input
stringlengths 4.25k
4.38k
| output
sequencelengths 1
1
| id
stringlengths 40
40
|
|---|---|---|
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: build.
Question: The contractor plans to build the deck soon, but the pool will have to wait. That's because the _ supplies will be here in a day.
Answer: deck.
Output:
| [
"The contractor plans to build the deck soon, but the pool will have to wait. That's because the _ supplies will be here in a month."
] | task034-85a7fd1b1f404c9cb31769d598664f10 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: book.
Question: James had been reading the pages of the book within the break period and he is yet to finish. The _ is long.
Answer: book.
Output:
| [
"James had been reading the pages of the book within the break period and he is yet to finish. The _ is short."
] | task034-9002ab9c6fb74db89533fb61937022d4 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: risk.
Question: There was more risk involved in leaving the phone in the car than the ring, because the _ was valuable.
Answer: phone.
Output:
| [
"There was more risk involved in leaving the phone in the car than the ring, because the _ was cheap."
] | task034-5e53ebd6235b45848caa0fef91bdfc4f |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: calories.
Question: She counted her calories for her diet and found she needed more so she ate a brownie instead of an apple since the _ has fewer.
Answer: apple.
Output:
| [
"She counted her calories for her diet and found she needed more so she ate a brownie instead of an apple since the _ has more."
] | task034-740e415a5b2e4d55a36e78d1b6508775 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: armpits.
Question: After showering, the woman shaved her armpits but not her legs because her _ were hairy.
Answer: armpits.
Output:
| [
"After showering, the woman shaved her armpits but not her legs because her _ were smooth."
] | task034-5ac9a353d1b942f7bd7642c302cd702d |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: bulb.
Question: James was scared of touching the light bulb because he though it will burn him like the candle light did. He realized the _ is cold.
Answer: bulb.
Output:
| [
"James was scared of touching the light bulb because he though it will burn him like the candle light did. He realized the _ is hot."
] | task034-b1041adeccbc492ebe69ddcd5e4c8998 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: fear.
Question: Colby was less fearful of swimming in the ocean than he was of swimming in the lake. The _ is cleaner.
Answer: ocean.
Output:
| [
"Colby was less fearful of swimming in the ocean than he was of swimming in the lake. The _ is dirtier."
] | task034-b54107e2a60f458180aafb08aa644588 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: friends.
Question: The introverted person does not make friends easily but the extroverted does because the _ is very shy.
Answer: introverted.
Output:
| [
"The introverted person does not make friends easily but the extroverted does because the _ is very confident."
] | task034-73684301b69d416cafaf7f6c59f37df6 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: settlement.
Question: We reached a settlement with the company but were roadblocked by the court. The _ was just more malleable in its dealings.
Answer: company.
Output:
| [
"We reached a settlement with the company but were roadblocked by the court. The _ was just more inflexible in its dealings."
] | task034-69ddf8d5f6ac43479e03f014a01af729 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: symptom.
Question: The symptom that I had this time was worse than the flu because the _ lasted three weeks.
Answer: symptom.
Output:
| [
"The symptom that I had this time was worse than the flu because the _ lasted three days."
] | task034-9f1709dbe1f543569fb5310f88f93290 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: bike.
Question: I tried to place the grocery bag in the basket on the bike, but the _ was too large.
Answer: bag.
Output:
| [
"I tried to place the grocery bag in the basket on the bike, but the _ was too small."
] | task034-97dd76f995a44be8abca91d624187a9c |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: fussy.
Question: The baby was fussy so instead of giving her a rattle, I gave her a pacifier because the _ keeps the baby awake.
Answer: rattle.
Output:
| [
"The baby was fussy so instead of giving her a rattle, I gave her a pacifier because the _ helps the baby sleep."
] | task034-6b124c66a23044818c83cce5571dddf1 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: paper.
Question: Susie tried to use a knife to cut through paper, but it didn't work. The _ was too dull.
Answer: knife.
Output:
| [
"Susie tried to use a knife to cut through paper, but it didn't work. The _ was too thick."
] | task034-0c42eb1fc35a43cd898e29bdae06fc40 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: party.
Question: The party had plenty of booze but no drugs, since the facility allowed the _ to be sold freely.
Answer: booze.
Output:
| [
"The party had plenty of booze but no drugs, since the facility forbade the _ to be sold freely."
] | task034-4d50ebb91362499c8279dd520cc075b8 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: milk.
Question: I'm on a diet so I've started using skim milk in my coffee instead of cream. The _ is too rich.
Answer: cream.
Output:
| [
"I'm on a diet so I've started using skim milk in my coffee instead of cream. The _ isn't too rich."
] | task034-02c5523999f64b13b3abd9c6fcd9e3c1 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: change.
Question: I had to change jobs so I sold my house and moved to another country. The _ was my new home.
Answer: country.
Output:
| [
"I had to change jobs so I sold my house and moved to another country. The _ was my old home."
] | task034-961109ec62d74abca1c8fa8c232827f1 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: juice.
Question: The bartender poured the juice from the blender into the cocktail glass until the _ was full.
Answer: glass.
Output:
| [
"The bartender poured the juice from the blender into the cocktail glass until the _ was empty."
] | task034-aeb2966c241b4bb3a352f3b94ce8776a |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: yoga.
Question: The yoga instructor to not use the ball but hold the pose since the _ was mediocre for the exercise.
Answer: ball.
Output:
| [
"The yoga instructor to not use the ball but hold the pose since the _ was excellent for the exercise."
] | task034-d33bedcb152d402da8fe63195b8ec09a |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: lying.
Question: Bob made a promise to his wife to stop lying anymore but the _ was too hard.
Answer: promise.
Output:
| [
"Bob made a promise to his wife to stop lying anymore but the _ was too easy."
] | task034-0160e09b703d45e381134cb8496e4bcf |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: vehicle.
Question: Mark prefers to lease a vehicle instead of rent it, because the _ is less expensive.
Answer: lease.
Output:
| [
"Mark prefers to lease a vehicle instead of rent it, because the _ is more expensive."
] | task034-9a756451a33f471a895b549a13f3715d |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: date.
Question: My date was at 5:00 and I needed to catch the train at 6:00, so there wasn't much time. Luckily, the _ was short, so it worked out.
Answer: date.
Output:
| [
"My date was at 5:00 and I needed to catch the train at 6:00, so there wasn't much time. Luckily, the _ was delayed, so it worked out."
] | task034-e845c13d61c04bf28c182df85b2faacb |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: piercing.
Question: The girl decided to get her ear pierced rather than her lip because she thought piercing the _ would hurt more.
Answer: lip.
Output:
| [
"The girl decided to get her ear pierced rather than her lip because she thought piercing the _ would hurt less."
] | task034-986702444a7040ceb6513fbcbc14bf1d |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: leggings.
Question: Jill dropped her ice cream onto her khakis and decided to put on jeans. She needed to put on different leggings because the _ are dirty.
Answer: khakis.
Output:
| [
"Jill dropped her ice cream onto her khakis and decided to put on jeans. She needed to put on different leggings because the _ are clean."
] | task034-ec454f9168f44dfc979cbd34648d11f4 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: violin.
Question: John played a violin at the concert instead of using his guitar because the _ was more modern.
Answer: guitar.
Output:
| [
"John played a violin at the concert instead of using his guitar because the _ was more classical."
] | task034-1ba90be1d9204067aabc1659b1a24831 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: seasoning.
Question: Sally liked to use seasoning when she was cooking meat on the grill. The _ was more flavorful.
Answer: meat.
Output:
| [
"Sally liked to use seasoning when she was cooking meat on the grill. The _ added flavor."
] | task034-e9d7783cfbe54f6f9e5b98a89ee1b512 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: potty.
Question: The man known as Jeff had to remove the dog's poop from the potty because the _ was clean.
Answer: potty.
Output:
| [
"The man known as Jeff had to remove the dog's poop from the potty because the _ was dirty."
] | task034-bc6806b4e1bd43f7b8af55a8fafbe20e |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: beans.
Question: The beans that I bought are better than the vegetables, because the _ are fresh.
Answer: beans.
Output:
| [
"The beans that I bought are better than the vegetables, because the _ are old."
] | task034-eaa2bcbcf3674537b8e624b541a25142 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: blocks.
Question: I was more annoyed with my kid's blocks than her dolls because the _ squeaked when I stepped on them.
Answer: dolls.
Output:
| [
"I was more annoyed with my kid's blocks than her dolls because the _ hurt when I stepped on them."
] | task034-8995bf4a7c344051b6ca5043a75caf28 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: fabric.
Question: He wanted to make a fabric gift box but didn't have thick batting to line it with only thin cotton. The _ was flimsy.
Answer: cotton.
Output:
| [
"He wanted to make a fabric gift box but didn't have thick batting to line it with only thin cotton. The _ was substantial."
] | task034-d84e8f7248d8466db16639aa22e0860d |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: nursery.
Question: The nursery did worse at it's downtown location versus being in the suburban area. The _ must have been burdensome to find.
Answer: downtown location.
Output:
| [
"The nursery did worse at it's downtown location versus being in the suburban area. The _ must have been convenient to find."
] | task034-ee081dad608c4ef99146b38da4ab71b1 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: options.
Question: When choosing between your options in Nice, it's better to visit the caves instead of the gardens because the _ are a lot less crowded.
Answer: caves.
Output:
| [
"When choosing between your options in Nice, it's better to visit the caves instead of the gardens because the _ are a lot more crowded."
] | task034-2c63f2c2cfaf456e9081bf9ca1c3886d |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: Work.
Question: Seth was running late for work, so he hopped on a train instead of his motorcycle. The _ was not fast enough.
Answer: motorcycle.
Output:
| [
"Seth was running late for work, so he hopped on a train instead of his motorcycle. The _ was much faster."
] | task034-9fe16409cf4644c2a8d22052cd377945 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: products.
Question: When Tim discovered some products missing, he accused the employees of giving too many samples. They said the _ were important for making sales.
Answer: samples.
Output:
| [
"When Tim discovered some products missing, he accused the employees of giving too many samples. He said the _ were important for making a profit."
] | task034-4ef8d7453ed14cd7b7f00c404946024d |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: eyes.
Question: John decided to stop wearing glasses and start using contacts instead, because the _ were weak for correcting his eyes.
Answer: glasses.
Output:
| [
"John decided to stop wearing glasses and start using contacts instead, because the _ were strong for correcting his eyes."
] | task034-940eb832f1e24bb48af18dae64b8a37e |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: slipped.
Question: John slipped while running on the track and they decided to use the lawn for the event. The _ is slippery.
Answer: track.
Output:
| [
"John slipped while running on the track and they decided to use the lawn for the event. The _ is dry."
] | task034-cf4a86b9eace44848f263f4b5b448822 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: recycle.
Question: The Northern council encouraged people to recycle but the Southern council did not because the _ council was very conscientious.
Answer: Northern.
Output:
| [
"The Northern council encouraged people to recycle but the Southern council did not because the _ council was very slapdash."
] | task034-e45809d4696543b591fa33eb873a9a3a |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: hair.
Question: Selena wanted the hairdresser to cut her hair with a razor instead of a clipper, because the _ always left rough edges.
Answer: clipper.
Output:
| [
"Selena wanted the hairdresser to cut her hair with a razor instead of a clipper, because the _ always gave an edgier look."
] | task034-f32f9c31bd224dcf8a75d31bba6a333c |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: learned.
Question: He had learned how to make art, but didn't know if it was his passion or a job. The _ seemed like a better outcome.
Answer: passion.
Output:
| [
"He had learned how to make art, but didn't know if it was his passion or a job. The _ seemed like a worse outcome."
] | task034-2efdc2c70742487883397571efcbb6fd |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: fee.
Question: The man wanted to use the money to pay the fee but the _ was too much.
Answer: fee.
Output:
| [
"The man wanted to use the money to pay the fee but the _ was too little."
] | task034-53047c30bb434ce99f88c83a35f59ec6 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: mushroom.
Question: The dinner was late because I didn't have any mushrooms for the sauce or to go with the noodles. The pasta needed a _ to go on it.
Answer: sauce.
Output:
| [
"The dinner was late because I didn't have any mushrooms for the sauce or to go with the noodles. The soup needed _ to go in it."
] | task034-18776c16d5cb45cd8dea6ffd4961d97b |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: morning.
Question: We avoided the park in the morning but visited the museum instead, as the _ was crowded at that time.
Answer: park.
Output:
| [
"We avoided the park in the morning but visited the museum instead, as the _ was roomy at that time."
] | task034-8338141b23214729b11c87bf4ade968f |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: car.
Question: Bobby couldn't get the new car he bought into the garage because the _ was to big.
Answer: car.
Output:
| [
"Bobby couldn't get the new car he bought into the garage because the _ was to small."
] | task034-462fa58b17e3455aad621df05d89129f |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: meltdowns.
Question: The kids meltdowns were a lot more intense than the child's tantrums because the _ were controllable.
Answer: meltdowns.
Output:
| [
"The kids meltdowns were a lot more intense than the child's tantrums because the _ were uncontrollable."
] | task034-d9eb4e74bf66493cb510949ccb08b0b9 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: fruit.
Question: The diet consisted mostly of fruit instead of things like meat, because the fact that the _ was alkaline.
Answer: fruit.
Output:
| [
"The diet consisted mostly of fruit instead of things like meat, because the fact that the _ was acidic."
] | task034-7856c79013894a8393fd6c39d22e16fe |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: apologize.
Question: The company had to apologize but the individual did not because the _ was held to be negligent.
Answer: company.
Output:
| [
"The company had to apologize but the individual did not because the _ was held to be blameless."
] | task034-0193a7029e4d4bcca070487e326d6988 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: planning.
Question: When planning a wedding it's important all plans and details are accounted for. The _ are less serious.
Answer: plans.
Output:
| [
"When planning a wedding it's important all plans and details are accounted for. The _ are more serious."
] | task034-0f8a788090eb4f518d2687e0a78d5863 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: measurement.
Question: James took the measurement of the pants and the sleeves and found that the _ were longer.
Answer: pants.
Output:
| [
"James took the measurement of the pants and the sleeves and found that the _ were shorter."
] | task034-8af186299b83444e96ac23bc570278d9 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: home.
Question: Nobody wants to go home but people do not mind attending the meeting at the hall. The _ is closer.
Answer: hall.
Output:
| [
"Nobody wants to go home but people do not mind attending the meeting at the hall. The _ is farther."
] | task034-218b48825da34503920152bc32924b5b |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: test.
Question: We moved out of the room to take the test in the hallway, as the _ made focusing difficult.
Answer: room.
Output:
| [
"We moved out of the room to take the test in the hallway, as the _ made focusing easier."
] | task034-f0944baaae44456790c8aab9e47a07ce |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: hair.
Question: Kim had to cut her hair either short or medium length, she decided on the _ length because the cooler was warmer.
Answer: medium.
Output:
| [
"Kim had to cut her hair either short or medium length, she decided on the _ length because it was more adventurous."
] | task034-90878e07986449a9b3a5f216cb6c861f |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: notes.
Question: The notes taken by James while reading is not enough to jot down on few pages because the _ are long.
Answer: notes.
Output:
| [
"The notes taken by James while reading is not enough to jot down on few pages because the _ are short."
] | task034-2bf0329e747340eba7605cebbbc700b6 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: uniform.
Question: The uniform previously used was out of style, the new dress code is more appropriate. The _ is trendy.
Answer: dress code.
Output:
| [
"The uniform previously used was out of style, the new dress code is more appropriate. The _ is outdated."
] | task034-fca1d9e1159b46c28aa249fb0627a45b |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: cutting board.
Question: She thought the cutting board was just right for the counter because the _ was bigger.
Answer: counter.
Output:
| [
"She thought the cutting board was just right for the counter because the _ was smaller."
] | task034-183f062d091046a382c9677ea045bb7d |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: insect.
Question: The yard had an insect problem of late so he purchased both bug spray and a bug zapper but the _ was too noxious.
Answer: bug spray.
Output:
| [
"The yard had an insect problem of late so he purchased both bug spray and a bug zapper but the _ was too loud."
] | task034-cf86579ff13743bf92bc05bc008dd91e |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: codfish.
Question: Ethan wanted to bake codfish for dinner, but he couldn't fit it into the pan because the _ was huge.
Answer: codfish.
Output:
| [
"Ethan wanted to bake codfish for dinner, but he couldn't fit it into the pan because the _ was small."
] | task034-1ca7cce7650d489797a59cc922eca168 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: cookie.
Question: She baked a single layer of cake and a giant cookie, the _ rose as it baked.
Answer: cake.
Output:
| [
"She baked a single layer of cake and a giant cookie, the _ flattened out as it baked."
] | task034-f6defb9eb44e4e8dae30ecce17749887 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: Date.
Question: Bob preferred the movie over dinner on the date because the _ didn't require conversation.
Answer: movie.
Output:
| [
"Bob preferred the movie over dinner on the date because the _ required more conversation."
] | task034-9ff51ed77d3249bcb48de3c754086699 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: grew.
Question: We grew the plants in the pot soil and not in the ground because the _ was soft.
Answer: pot soil.
Output:
| [
"We grew the plants in the pot soil and not in the ground because the _ was hard."
] | task034-80ff0023e93c46a0a0357b7d6483bde2 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: landscaping.
Question: The landscaping looked better at the house than at the cottage because the _ owners used a hardworking company.
Answer: house.
Output:
| [
"The landscaping looked better at the house than at the cottage because the _ owners used a lazy company."
] | task034-cea42a2b01264150b205f457dac92e3e |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: coconut.
Question: For your Hawaiian themed party, I think you should serve coconut instead of grapes since the _ is more of a tropical fruit.
Answer: coconut.
Output:
| [
"For your Hawaiian themed party, I think you should serve coconut instead of grapes since the _ is less of a tropical fruit."
] | task034-1888434beae745ed8ffb6322a18b5e00 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: noise.
Question: The noise from the band was too loud for the speaker because the _ was intense.
Answer: noise.
Output:
| [
"The noise from the band was too loud for the speaker because the _ was cheap."
] | task034-1dc51af7cdd548079da29979232046d4 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: grams.
Question: The bowling ball was more grams than the basketball because the _ took up more mass.
Answer: bowling ball.
Output:
| [
"The bowling ball was more grams than the basketball because the _ took up less mass."
] | task034-a4253a21cb2b424faf60d83f2a93325b |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: disaster.
Question: After the disaster in the dark, we sold our car and bought a truck since the _ 's lights were dim.
Answer: car.
Output:
| [
"After the disaster in the dark, we sold our car and bought a truck since the _ 's lights were bright."
] | task034-f22069af77884eff909a553a7a540a3f |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: allergic.
Question: There was much more concern about being allergic to the pollen fragments and almost no more concern about the peanuts. So that meant that the _ were worrisome.
Answer: pollen fragments.
Output:
| [
"There was much more concern about being allergic to the pollen fragments and almost no more about the peanuts. So that meant that the _ were reassuring."
] | task034-865fcbc01ddd4afa8e535f4dee4c82ff |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: world.
Question: In their world fire was tolerated but not ice since the _ was small in number.
Answer: fire.
Output:
| [
"In their world fire was tolerated but not ice since the _ was large in number."
] | task034-2330f733a0cc45f7aecfa5cbc1935258 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: yarn.
Question: The cardigan took longer to knit than the sweater because the _ was knitted with finer yarn.
Answer: cardigan.
Output:
| [
"The cardigan took longer to knit than the sweater because the _ was knitted with chunkier yarn."
] | task034-27c6ae79800046a382169aa2c6e95213 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: party.
Question: Wendy wanted to melt chocolate onto the strawberries for the party, but the _ were too small.
Answer: strawberries.
Output:
| [
"Wendy wanted to melt chocolate onto the strawberries for the party, but the _ was too thick."
] | task034-a0cf5d61972b43ef9e6ef8e81090c087 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: thoughts.
Question: I loved having good thoughts rather than nightmares because the _ always made me happy.
Answer: thoughts.
Output:
| [
"I loved having good thoughts rather than nightmares because the _ always made me scared."
] | task034-428ee9bf9f8c40699dea4dcafede2e5c |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: night shift.
Question: The man had to drink a lot of coffee to stay alert and awake during the night shift, because the _ was boring.
Answer: shift.
Output:
| [
"The man had to drink a lot of coffee to stay alert and awake during the night shift, because the _ was stimulating."
] | task034-90bd3f1223de4a189c42a6e6f561b5cd |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: dreams.
Question: Jen made a list and a budget to help herself achieve her dreams, but the _ was easier for her to read.
Answer: list.
Output:
| [
"Jen made a list and a budget to help herself achieve her dreams, but the _ was harder for her to keep."
] | task034-2b6b6ccb3c584ce8bd654b41a098da7e |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: throat.
Question: The knife punctured the throat while deflecting off the bone, since the _ was softer.
Answer: throat.
Output:
| [
"The knife punctured the throat while deflecting off the bone, since the _ was harder."
] | task034-9e486f36df6a460b99b5d6a11dec9abe |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: incense.
Question: When burning either cone or stick incense, be aware that the _ incense burns smokier.
Answer: cone.
Output:
| [
"When burning either cone or stick incense, be aware that the _ incense burns clearer."
] | task034-ac43be23645944a790512fa9a95b213e |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: bakery.
Question: There was more donuts cooked in the morning at the bakery than at night because no one ate donuts at _ .
Answer: night.
Output:
| [
"There was more donuts cooked in the morning at the bakery than at night because everyone ate donuts in the _ ."
] | task034-584ce29e7b4c41d5b6f6684e9fc39188 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: phone.
Question: Janine wanted to get rid of her phone forever so she threw it in the lake instead of the fountain, because the _ was too shallow.
Answer: lake.
Output:
| [
"Janine wanted to get rid of her phone forever so she threw it in the lake instead of the fountain, because the _ was much deeper."
] | task034-7c67f5dff6954e6fba795796edf79524 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: tubes.
Question: The worker tried to store the tubes in the small boxes but the _ were too small.
Answer: boxes.
Output:
| [
"The worker tried to store the tubes in the small boxes but the _ were too large."
] | task034-fedc41ac724142b6b17607b617481415 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: home closing.
Question: Kim filled out a form for time off to prepare for the home closing on Friday and the _ was so easy.
Answer: form.
Output:
| [
"Kim filled out a form for time off to prepare for the home closing on Friday and the _ was so complicated."
] | task034-8900001c1dde48e9ba7dfff8ae977ea5 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: diet.
Question: Bob ate a diet of pasta and not meat before the marathon because the _ was less useful during the race.
Answer: meat.
Output:
| [
"Bob ate a diet of pasta and not meat before the marathon because the _ was more useful during the race."
] | task034-923f859186c54f2ebf6c805906dab4a3 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: fish bowl.
Question: The betta's fish bowl needed cleaned and the water replaced, since there was a lot of algae growing in it. The _ was removed after the cleaning.
Answer: algae.
Output:
| [
"The betta's fish bowl needed cleaned and the water replaced, since there was a lot of algae growing in it. The _ was removed and replaced after the cleaning."
] | task034-884707a69f1c4c4dafaae14f38f2e6e1 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: host.
Question: The host of the restaurant was offering samples of the bread instead of the fruit, since the _ was popular.
Answer: bread.
Output:
| [
"The host of the restaurant was offering samples of the bread instead of the fruit, since the _ was disliked."
] | task034-aa037a9d1da84bd1bbd07cbe1a9d6bb4 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: fork.
Question: The mother was choosing between the fork and the spoon, but the child was worse with the _ when eating.
Answer: spoon.
Output:
| [
"The mother was choosing between the fork and the spoon, but the child was better with the _ when eating."
] | task034-bea95bdc50df46b7ac03b560350607ba |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: bites.
Question: I was petrified to find if the bites were from fleas or bedbugs. I was relieved to find they were from the _ .
Answer: fleas.
Output:
| [
"I was petrified to find if the bites were from fleas or bedbugs. I was horrified to find they were from the _ ."
] | task034-21c3b9f56cdd46b5a260792719b0c164 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: backpack.
Question: Mary had used a tote bag more than a backpack, even though the _ is tinier.
Answer: tote bag.
Output:
| [
"Mary had used a tote bag more than a backpack, even though the _ is roomier."
] | task034-b33ef6ba7f174471bba624ac4fdee180 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: store.
Question: We could sell the watches at the store but not the clocks , since the _ were good for business.
Answer: watches.
Output:
| [
"We could sell the watches at the store but not the clocks , since the _ were bad for business."
] | task034-303206dc95e34365a201fcda1813689c |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: color.
Question: Amy prefers to wear bright color clothes over black clothes, because the _ are more suitable for the summer.
Answer: bright color clothes.
Output:
| [
"Amy prefers to wear bright color clothes over black clothes, because the _ are more suitable for the winter."
] | task034-a4a891e9d8954221829b93ff75a8606f |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: shallot.
Question: I returned the dish with the shallot but kept the one with the onion because the _ was cooked wrong.
Answer: shallot.
Output:
| [
"I returned the dish with the shallot but kept the one with the onion because the _ was cooked correctly."
] | task034-6f2556d22c354c2fac241eb50e37a65d |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: tea.
Question: My mom bought me tea that was worse than the coffee, because the _ was bad quality.
Answer: tea.
Output:
| [
"My mom bought me tea that was worse than the coffee, because the _ was good quality."
] | task034-54118e3c73354c1795cc9211b9c23239 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: Leash.
Question: Mike tried to put the leash on this dog around it's neck, but the _ was too small.
Answer: leash.
Output:
| [
"Mike tried to put the leash on this dog around it's neck, but the _ was too big."
] | task034-af4c9a15766f4b349eecaede3fc8f313 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: landscaping.
Question: The cottage looked better than the house because the landscaping of the _ was beautiful.
Answer: cottage.
Output:
| [
"The cottage looked better than the house because the landscaping of the _ was ugly."
] | task034-3f6f0b195a7543339c200eccc9b36794 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: ginger.
Question: It is a lot easier to consume ginger than turmeric because the _ is spicier.
Answer: turmeric.
Output:
| [
"It is a lot easier to consume ginger than turmeric because the _ is milder."
] | task034-2c9c1cdb11d949d091a73082099129de |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: milk.
Question: After taking a sip and burning his mouth, the man added milk to the tea because the _ was cold.
Answer: milk.
Output:
| [
"After taking a sip and burning his mouth, the man added milk to the tea because the _ was hot."
] | task034-0587fca9c7f24be1b7be6866f1e3f25b |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: time.
Question: John needed more time to complete reading the book. It turn out that the _ given to him is long.
Answer: book.
Output:
| [
"John needed more time to complete reading the book. It turn out that the _ given to him is short."
] | task034-3e1c261e660746d8b0ee7286a0a36820 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: theatre.
Question: The theatre wouldn't screen the film, but showed the cartoon instead, as the _ was a loss maker.
Answer: film.
Output:
| [
"The theatre wouldn't screen the film, but showed the cartoon instead, as the _ was a money maker."
] | task034-351ef30190204c47927c8618dfb87f46 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: bidet.
Question: The hotel luxury suite bathroom had both a bidet and a toilet for their guests, because the _ was European.
Answer: hotel.
Output:
| [
"The hotel luxury suite bathroom had both a bidet and a toilet for their guests, because the _ conserved water."
] | task034-8bad89d1433f46da84e2bd4ccf732fb4 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: symptoms.
Question: The medicine did not help with the patient's symptoms because the _ were too severe.
Answer: symptoms.
Output:
| [
"The medicine did not help with the patient's symptoms because the _ medicine was too old."
] | task034-180b5ff1e9424983a8e923c1061882ba |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: puppy.
Question: Upon returning home, Mary saw her shoes were safe, but the puppy had chewed up her curtains, because the _ were hard to reach.
Answer: shoes.
Output:
| [
"Upon returning home, Mary saw her shoes were safe, but the puppy had chewed up her curtains, because the _ were easy to reach."
] | task034-d5ddf15ff90648078ca83e3a2084589f |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: cold.
Question: She drank pill with sugar water then gargled with salt water because she caught a cold, the _ was ineffective for sore throat.
Answer: sugar water.
Output:
| [
"She drank pill with sugar water then gargled with salt water because she caught a cold, the _ was effective for sore throat."
] | task034-86489768b2f8444b978f07b0f4bca661 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: bedroom.
Question: The bedroom had to have the bureau replaced by the new closet , since the _ was too packed for all the new clothes.
Answer: bureau.
Output:
| [
"The bedroom had to have the bureau replaced by the new closet , since the _ was very roomy for all the new clothes."
] | task034-d52e6d56d2ab45d4bc8fff733ac18aee |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: love.
Question: Giving a shirt as a gift is a better token of love than giving a hat, as the _ is a stronger symbol of affection.
Answer: shirt.
Output:
| [
"Giving a shirt as a gift is a better token of love than giving a hat, as the _ is a lamer symbol of affection."
] | task034-a16ee2b397a746919df3d16ee6a37d7c |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: mail.
Question: John got his mail delivered to the church faster than the office because the _ is closer.
Answer: church.
Output:
| [
"John got his mail delivered to the church faster than the office because the _ is farther."
] | task034-7d7634d162924366a5ad36021b4b5016 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: how to tune a room.
Question: Millie tried to watch a video on the best on how to tune a room but the _ was too soft.
Answer: video.
Output:
| [
"Millie tried to watch a video on the best on how to tune a room but the _ was too loud."
] | task034-edb92d9f5c77498a91aa24254c458b43 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.