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: address.
Question: She had to address the issue with her children but had to choose between discussing or punishing them. The _ choice made her more uncomfortable.
Answer: discussing.
Output:
| [
"She had to address the issue with her children but had to choose between discussing or punishing them. The _ choice made her more comfortable."
] | task034-0bf17ba4d1184fa79aaf3aff06c9031c |
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: table.
Question: John cleaned the dusty table with a damp towel and the _ became very clean.
Answer: table.
Output:
| [
"John cleaned the dusty table with a damp towel and the _ became very dirty."
] | task034-5c99cb5c76d04dcf849b9b329410adb8 |
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 mexican drive-thru sold out of chips but not beans, since _ were unhealthy and more popular.
Answer: chips.
Output:
| [
"The mexican drive-thru sold out of chips but not beans, since _ were healthy and less popular."
] | task034-eb1ac5a1853d45d8bb922fd53bcec1c2 |
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: card.
Question: The Christmas card was beautiful but the Easter card was not as the artist who did the _ card was very talented.
Answer: Christmas.
Output:
| [
"The Christmas card was beautiful but the Easter card was not as the artist who did the _ card was very inept."
] | task034-d46327e353fd46b1b26898e9bdc98504 |
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: adrenaline rush.
Question: He loves the adrenaline rush so he bought the sports car over the family car, because the _ is faster.
Answer: sports car.
Output:
| [
"He loves the adrenaline rush so he bought the sports car over the family car, because the _ is slower."
] | task034-7019a081bc8c4d82b7e6ca0890e39bb4 |
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: cake.
Question: I took the cake out of the oven early and let the pie cook the full time. The _ was raw.
Answer: cake.
Output:
| [
"I took the cake out of the oven early and let the pie cook the full time. The _ was done."
] | task034-a319775b3d504b4886082f9ab36c64be |
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: company.
Question: The company paid a higher salary than the business because the _ made more money.
Answer: company.
Output:
| [
"The company paid a lower salary than the business because the _ made more money."
] | task034-a9a38319913c4ed3b2aa722b6c034e6a |
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: yard sale.
Question: While shopping at the yard sale, the woman passed on the dishes, but bought the spoons, because she already had many of the _ .
Answer: dishes.
Output:
| [
"While shopping at the yard sale, the woman passed on the dishes, but bought the spoons, because she needed more of the _ ."
] | task034-5d8a1eeafefc437da85c92b2bad7a53d |
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: School.
Question: Adam dreaded returning to school because his backpack was ripped and his mom didn't buy him a new one. The _ needed replacement.
Answer: backpack.
Output:
| [
"Adam dreaded returning to school because his backpack was ripped and his mom didn't buy him a new one. The _ instilled fear."
] | task034-f2cb65d691c8461ea8196d44548ae67c |
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: vacation.
Question: The vacation was more expensive than the business trip was because the _ was longer.
Answer: vacation.
Output:
| [
"The vacation was more expensive than the business trip was because the _ was shorter."
] | task034-8ddb1de71edf4032acf0da6122306fbd |
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: park.
Question: The experienced pilot was unable to park the commercial airplane in the hangar because the _ was too wide.
Answer: plane.
Output:
| [
"The experienced pilot was unable to park the commercial airplane in the hangar because the _ was too narrow."
] | task034-4d56042334f940b0a6dcc1f99a752270 |
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: I went to get some fruit off the tree in the backyard, but the _ was bare.
Answer: tree.
Output:
| [
"I went to get some fruit off the tree in the backyard, and the _ was ripe."
] | task034-86ed54c22727499bad6841b42543dabe |
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: traveling.
Question: When I am traveling, I carry pro-grade prime lenses compared to pro-grade zoom lenses because the _ lenses are heavy.
Answer: pro-grade zoom.
Output:
| [
"When I am traveling, I carry pro-grade prime lenses compared to pro-grade zoom lenses because the _ lenses are light."
] | task034-b4dad57874ff4bd0b36f9effa5192f2f |
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: drain pipe.
Question: The wedding ring fell into the drain pipe, but the necklace didn't fall. The _ was lost.
Answer: ring.
Output:
| [
"The wedding ring fell into the drain pipe, but the necklace didn't fall. The _ was safe."
] | task034-700940123a054f98a0ddc830306400fd |
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: install.
Question: The app was easier to install on the computer than the phone because there was more free space on the _ .
Answer: computer.
Output:
| [
"The app was harder to install on the computer than the phone because there was more free space on the _ ."
] | task034-58f0e4cd8c4e4982b1ed3d13e83d662f |
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: bruising.
Question: The bruising was prevalent on his face but not his neck, as the majority of blows landed on the _ .
Answer: face.
Output:
| [
"The bruising was prevalent on his face but not his neck, as the minority of blows landed on the _ ."
] | task034-33a5d08a9c0e4beda2fbacc29b0df697 |
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 symptoms appeared in the nose now, not the mouth, as the _ was now infected.
Answer: nose.
Output:
| [
"The symptoms appeared in the nose now, not the mouth, as the _ was now healed."
] | task034-a192986b9ecc43a5889fa8b0636b8fe1 |
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: chest.
Question: Karen could not fit the antique chest she just bought in her car's trunk because the _ was too small.
Answer: trunk.
Output:
| [
"Karen could not fit the antique chest she just bought in her car's trunk because the _ was too big."
] | task034-84f70d300d264c61a726cbff3975b01e |
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: vodka.
Question: The vodka that we were going to be consuming later in the evening was a different color than the whiskey, and the _ was lighter.
Answer: vodka.
Output:
| [
"The vodka that we were going to be consuming later in the evening was a different color than the whiskey, and the _ was darker."
] | task034-36613c81f6824ad7b43f16fe9d5e9112 |
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: eating.
Question: Amy was eating apples for the week but had to throw away all the oranges because the _ were disgusting.
Answer: oranges.
Output:
| [
"Amy was eating apples for the week but had to throw away all the oranges because the _ were delicious."
] | task034-add35b19067540269a9fce9d38fa196b |
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: parrot.
Question: The cat was harder to take care of than the parrot because the _ ate more food.
Answer: cat.
Output:
| [
"The cat was harder to take care of than the parrot because the _ ate less food."
] | task034-c1e6a19fa61d408780af9a7928498308 |
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: student.
Question: The student sat at a desk, but the teacher moved him to a table because there was more room at the _ .
Answer: table.
Output:
| [
"The student sat at a desk, but the teacher moved him to a table because there was less room at the _ ."
] | task034-d620fe33b0e349ceae7147acfc068ac6 |
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: basketball.
Question: The sac has more space for one more basketball. The _ is a big one.
Answer: sac.
Output:
| [
"The sac has more space for one more basketball. The _ is a small one."
] | task034-426dcdf246d94e81876962ed54f11143 |
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: treats.
Question: One way to make rice crispy treats will need chocolate chips and marshmallows with the _ starting off hard.
Answer: chocolate.
Output:
| [
"One way to make rice crispy treats will need chocolate and marshmallows with the _ starting off soft."
] | task034-d5de02f52eb94d6a9ae3b47ad997fc73 |
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: battery.
Question: The battery of the the phone died faster than the laptop battery, because the _ was always on.
Answer: phone.
Output:
| [
"The battery of the the phone died faster than the laptop battery, because the _ was always off."
] | task034-be63ded151254a83a989397513e326d6 |
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: teaching.
Question: The professor liked teaching using picture cards instead of models, because the _ don't appeal to their students.
Answer: models.
Output:
| [
"The professor liked teaching using picture cards instead of models, because the _ engaged their students."
] | task034-b5cffbba57604aed8f17c9489837f1e4 |
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: rules.
Question: The rules of seven card stud were tougher to understand than five card draw. _ is the more complex game.
Answer: seven card stud.
Output:
| [
"The rules of seven card stud were tougher to understand than five card draw. _ is the more simple game."
] | task034-be147cc258464a458c708a4992abcb83 |
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: Bananas.
Question: Missy liked bananas a lot and cherries less, so she was mad that her sundae had few of the _ on it.
Answer: Bananas.
Output:
| [
"Missy liked bananas a lot and cherries less, so she was mad that her sundae had many of the _ on it."
] | task034-3605f093be374328ad8348a2fc78ff2c |
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: evacuation.
Question: The people in the building were moved to the hospital after the evacuation, as the _ was dangerous for them.
Answer: building.
Output:
| [
"The people in the building were moved to the hospital after the evacuation, as the _ was safe for them."
] | task034-4facc8c9bb7147978cbeee4f8688ee53 |
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: abortion.
Question: The abortion was a lot more difficult than her last pregnancy, because the _ gave her guilt.
Answer: abortion.
Output:
| [
"The abortion was a lot more difficult than her last pregnancy, because the _ gave her joy."
] | task034-515cea2006524e919f83c5fd21f7a4f6 |
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: ski.
Question: The ski resort had new rules, so we had to use skis instead of snowboards as the _ were safer.
Answer: skis.
Output:
| [
"The ski resort had new rules, so we had to use skis instead of snowboards as the _ were more dangerous."
] | task034-b254e4a2e062429086ec4d9c664af27a |
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: mouse.
Question: The mouse got stuck in a hole while running with a wood in its mouth. The _ is wide.
Answer: wood.
Output:
| [
"The mouse got stuck in a hole while running with a wood in its mouth. The _ is small."
] | task034-d944c13e726147aebef2513d9acff149 |
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: rabbit.
Question: John replaced lettuce in his rabbit's diet with carrot, because the _ was unhealthy for the rabbit.
Answer: lettuce.
Output:
| [
"John replaced lettuce in his rabbit's diet with carrot, because the _ was wonderful for the rabbit."
] | task034-7ff5dce54d944cdaaab96ccbfb210332 |
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: walker.
Question: The senior's walker fit in the hallway but got stuck in the doorway because the _ was wider .
Answer: hallway.
Output:
| [
"The senior's walker fit in the hallway but got stuck in the doorway because the _ was narrower ."
] | task034-fed0e050d2024d21896f6496d15a0d81 |
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: rain.
Question: Alicia was outside in the grass when the rain began to start, and she ended up getting dirty because the _ was heavy.
Answer: rain.
Output:
| [
"Alicia was outside in the grass when the rain began to start, and she ended up getting dirty because the _ was wet."
] | task034-c46545425c2e48a9b2db513dbb134154 |
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: report.
Question: It was her responsibility to create the report instead before the paper, because the _ was detailed.
Answer: report.
Output:
| [
"It was her responsibility to create the report instead before the paper, because the _ was simple."
] | task034-9b50dad49a98473f98f3fdebaa971e09 |
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: job.
Question: Arnold applied for a job at the factory. When it was offered, her rejected it because the _ paid too little.
Answer: job.
Output:
| [
"Arnold applied for a job at the factory. When it was offered, her rejected it because the _ was too far away."
] | task034-341bdcabe0954701ab1cae2f61baa7d3 |
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: rice.
Question: The rice needed more seasoning, but the pasta tasted just fine. The _ must have been bland.
Answer: rice.
Output:
| [
"The rice needed more seasoning, but the pasta tasted just fine. The _ must have been spicy."
] | task034-b71a0dfaf6784d41aede17fbc3e39613 |
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: muscle.
Question: The rod failed to puncture him, but the arrow succeeded. The _ was weak against his muscle.
Answer: rod.
Output:
| [
"The rod failed to puncture him, but the arrow succeeded. The _ was strong against his muscle."
] | task034-93702ca419dc45c9b6756efcdb7a2b3a |
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: socks.
Question: I could't decide what type of socks to wear, business or casual. When my boss said the meeting was formal, I picked _ .
Answer: business.
Output:
| [
"I could't decide what type of socks to wear, business or casual. When my boss said the meeting was laid back, I picked _ ."
] | task034-339edf39e2354e04887505a9e8341b9b |
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: beneficial.
Question: The syrup was more beneficial than the medicine for suppressing the cough, since the _ was expired.
Answer: medicine.
Output:
| [
"The syrup was more beneficial than the medicine for suppressing the cough, since the _ was new."
] | task034-1edfbd093aca4579947f8af771a5a6d5 |
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: sausage.
Question: Sausage and bacon are both good meats, but the _ is more versatile as it can be used in more soups and casseroles.
Answer: sausage.
Output:
| [
"Sausage and bacon are both good meats, but the _ is less versatile as it can be used in less soups and casseroles."
] | task034-84e0e77a559b488db094871f18b883d8 |
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: computers.
Question: The computers in the office ran faster than the ones in the school because the ones in the _ were older.
Answer: school.
Output:
| [
"The computers in the office ran faster than the ones in the school because the ones in the _ were newer."
] | task034-6ff41312759b40bab55976cc7f84b358 |
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: warned.
Question: The girl was warned not to touch the pan, but told it was ok to touch the pot, because the _ had been heated.
Answer: pan.
Output:
| [
"The girl was warned not to touch the pan, but told it was ok to touch the pot, because the _ pot had cooled."
] | task034-c366cc71ffb34909b517d0d06abfc114 |
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: Letter.
Question: Sherry sent an email to her boss instead of a letter because the _ would be harder for her to compose.
Answer: letter.
Output:
| [
"Sherry sent an email to her boss instead of a letter because the _ would be easier for her to compose."
] | task034-5733627416574d18b1a10c9888e1a225 |
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: Sandwich.
Question: Trisha is happy to swap her sandwich for my salad. She thinks the _ fattening for her.
Answer: sandwich.
Output:
| [
"Trisha is happy to swap her sandwich for my salad. She thinks the _ healthy for her."
] | task034-304325f4f34a48cf8ac212bbee487a94 |
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: towel.
Question: Sam's towel was always wet because he forgot to hang it after he took a shower. When the _ was hung, the towel finally dried.
Answer: towel.
Output:
| [
"Sam's towel was always wet because he forgot to hang it after he took a shower. When the _ was broken, the towel finally dried."
] | task034-84735efbda4c47219fa823b5d4641150 |
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: memory card.
Question: You can store more JPEG files in a given memory card compared to RAW files because size of the _ files are big.
Answer: RAW.
Output:
| [
"You can store more JPEG files in a given memory card compared to RAW files because size of the _ files are small."
] | task034-972c911fc3eb49cfbc8b18dc8ea49471 |
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: rake.
Question: One can pick up more the leaves than the gravels with a rake because the _ are tiny.
Answer: gravels.
Output:
| [
"One can pick up more the leaves than the gravels with a rake because the _ are large."
] | task034-568defe80d27491d8a8fcc35ab5a361d |
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: pumpkin.
Question: The woman was able to fit the pumpkin in the cart, even though the _ was very large.
Answer: pumpkin.
Output:
| [
"The woman was able to fit the pumpkin in the cart, even though the _ was very small."
] | task034-242a37d1fe034e1eb552dc87e72ad18c |
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: Caffeine.
Question: The drink was rich in caffeine and low in sugar because the _ was suffice in keeping the drinker awake.
Answer: caffeine.
Output:
| [
"The drink was rich in caffeine and low in sugar because the _ was not enough to keep the drinker awake."
] | task034-39736c7eba3c423490ffcc2d1bf5ae0c |
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: cheese.
Question: The chef preferred to use the cream cheese instead of the blue because the _ cheese had a mild flavor.
Answer: cream.
Output:
| [
"The chef preferred to use the cream cheese instead of the blue because the _ cheese had a pungent flavor."
] | task034-57789ded9309482e9e8b2f885fd9bf28 |
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: friendships.
Question: The local cafe was a much better place for making new friendships than the library was, as the _ was accommodating to social activity.
Answer: cafe.
Output:
| [
"The local cafe was a much better place for making new friendships than the library was, as the _ was hostile to social activity."
] | task034-4ef7a7c5966f4d57812a472ee4246ff3 |
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: addiction.
Question: His addiction was getting bad so they handed him over to therapy and stopped taking medicine since the _ was cheaper.
Answer: therapy.
Output:
| [
"His addiction was getting bad so they handed him over to therapy and stopped taking medicine since the _ was more expensive."
] | task034-70dfc8fb37914096956d47543c1b2a75 |
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: kitchen.
Question: My mother accidentally burned the meatloaf that we had planned to have for dinner in the kitchen oven, because the _ was too dry.
Answer: meatloaf.
Output:
| [
"My mother accidentally burned the meatloaf that we had planned to have for dinner in the kitchen oven, because the _ was too hot."
] | task034-c342fb7152d54ba99106d4d1b862f19e |
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: clean.
Question: I had to clean the carpets and I ended up replacing them with tiles. The _ looked old.
Answer: carpets.
Output:
| [
"I had to clean the carpets and I ended up replacing them with tiles. The _ looked shiny."
] | task034-63070d3df2a24322bea477e5060bda72 |
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: trees.
Question: The bushes look a lot worse than the trees in front of the house because the _ are less taken care of.
Answer: bushes.
Output:
| [
"The bushes look a lot worse than the trees in front of the house because the _ are more taken care of."
] | task034-ab9a3fc8e1414ae68fdec8a387a0fd0f |
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: necklace.
Question: We sold the necklace on the internet instead of at a store, because the _ was more secure for all the money involved.
Answer: internet.
Output:
| [
"We sold the necklace on the internet instead of at a store, because the _ was more vulnerable for all the money involved."
] | task034-0b0e8e4572a642bb9828aeb70f03a5f0 |
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: short.
Question: The short do not get wet even though James swam with it but the shirt is dripping with water. The _ is absorbent.
Answer: shirt.
Output:
| [
"The short do not get wet even though James swam with it but the shirt is dripping with water. The _ is waterproof."
] | task034-5ba29e911d114985bc9ab750719371cc |
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: conversation.
Question: Alice enjoyed conversation with philosophers and not salesmen because _ liked to talk about abstract topics.
Answer: philosophers.
Output:
| [
"Alice enjoyed conversation with philosophers and not salesmen because _ liked to talk about concrete topics."
] | task034-a5bf03576bd54987a827d064851a2e2f |
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: package.
Question: The package with the books was heavier than the one with the clothing so it costed more to send the _ .
Answer: books.
Output:
| [
"The package with the books was lighter than the one with the clothing so it costed more to send the _ ."
] | task034-6df2d8ce9808463494c947aa67fed4b7 |
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: splinter.
Question: The metal splinter was easier to remove than the last time I got a wood one. The _ must have an easier time coming out.
Answer: metal.
Output:
| [
"The metal splinter was easier to remove than the last time I got a wood one. The _ must have a tougher time coming out."
] | task034-6772ca4c844c476e90892c3d4d62811a |
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: cost.
Question: I didn't like the cost of the game compared to the price of the model, because the _ was so expensive.
Answer: game.
Output:
| [
"I didn't like the cost of the game compared to the price of the model, because the _ was so cheap."
] | task034-81a15b4382694f8aa65738a269779397 |
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 nose piercing Sarah got was infected but the tattoo healed well. The _ was swollen.
Answer: nose.
Output:
| [
"The nose piercing Sarah got was infected but the tattoo healed well. The _ was healed."
] | task034-8bb82a6d18f94561bb141e974eb861ca |
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: spider.
Question: Kim had a pet spider she wanted to put in the terrarium with her climbing rock but the _ was too large.
Answer: terrarium.
Output:
| [
"Kim had a pet spider she wanted to put in the terrarium with her climbing rock but the _ was too big."
] | task034-631627cb730642b7a81d5ed33c21fcdd |
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: discipline.
Question: Chemistry was a less popular discipline at the school than biology, as the _ was less lucrative as a field.
Answer: chemistry.
Output:
| [
"Chemistry was a less popular discipline at the school than biology, as the _ was more lucrative as a field."
] | task034-a7f97b3b11d14d86a541bb5bec561410 |
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: firearm.
Question: Mary needed more training to use the rifle than the firearm, because she was inexperienced using the _ .
Answer: rifle.
Output:
| [
"Mary needed more training to use the rifle than the firearm, because she was experienced using the _ ."
] | task034-29d32139c24b4bfd981e205bcbebf8f6 |
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: light.
Question: In the evening, I turned on the light in the room because the _ was dark.
Answer: room.
Output:
| [
"In the evening, I turned on the light in the room because the _ were bright."
] | task034-12af3b3ecbc0408b8b763146b6733997 |
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: celebrate.
Question: As part of a Jewish family, we celebrate Hanukkah instead of Christmas as the _ is familiar to us.
Answer: Hanukkah.
Output:
| [
"As part of a Jewish family, we celebrate Hanukkah instead of Christmas as the _ is strange to us."
] | task034-ad23bbe565cc4cb6a46078c0100a9873 |
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: clothes.
Question: The housekeeper tried to fit the clothes on the chairs but the _ were too large.
Answer: clothes.
Output:
| [
"The housekeeper tried to fit the clothes on the chairs but the _ were too small."
] | task034-7cdf7f1f1be84672b92c2fa406bf2ef2 |
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: grilled.
Question: Since they grilled hot dogs, she chose mango salsa as an alternative to ketchup, since the _ was uncommon.
Answer: mango salsa.
Output:
| [
"Since they grilled hot dogs, she chose mango salsa as an alternative to ketchup, since the _ was ordinary."
] | task034-a22d780dbedd4304b61b23f5555e24f7 |
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: boyfriend.
Question: Jane's boyfriend bought her a bracelet instead of a necklace for her birthday because _ was her preferred jewelry.
Answer: bracelet.
Output:
| [
"Jane's boyfriend bought her a bracelet instead of a necklace for her birthday because _ was her undesired jewelry."
] | task034-a3cb46d112014cfe95d1e86a160670d4 |
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: strap.
Question: The barber used the strap to sharpen the razors but not the scissors, since the _ worked better after sharpening.
Answer: razors.
Output:
| [
"The barber used the strap to sharpen the razors but not the scissors, since the _ worked worse after sharpening."
] | task034-c5252ccac7c2406fa00bb794fc6c09e6 |
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: homework.
Question: For my homework I have to interview either my mom or my dad. I chose the _ because he has great stories.
Answer: dad.
Output:
| [
"For my homework I have to interview either my mom or my dad. I chose the _ because she has a great job."
] | task034-8253122878364eaa8a0e426c67605bd4 |
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: Rolex.
Question: James wanted a Rolex but could only afford a plastic watch, and he was disappointed in the quality of the _ .
Answer: plastic watch.
Output:
| [
"James wanted a Rolex but could only afford a plastic watch, and he was disappointed in the price of the _ ."
] | task034-edcd70b25c274b1f81b44a61ab96deba |
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: remote control.
Question: The remote control worked better than the toy car because the batteries in the _ were fresher.
Answer: remote control.
Output:
| [
"The remote control worked better than the toy car because the batteries in the _ were older."
] | task034-aa416dff398e46699568c6a9d5800366 |
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: chocolate.
Question: She loved chocolate so much that her body was found dead and bloated in a pile of the _ .
Answer: chocolate.
Output:
| [
"She loved chocolate so much that her body was found dead and bloated in a pile of the _ smothered."
] | task034-1e86369cd1514faca59d46cd1a8ed9c9 |
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: trim.
Question: I decided to go to the barbershop instead of my usual salon because I just needed a trim and the _ was faster and cheaper.
Answer: barbershop.
Output:
| [
"I decided to go to the barbershop instead of my usual salon because I just needed a trim and the _ was slower and expensive."
] | task034-5fc7fc6ecfdb4efd85bdf59905633d66 |
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: impress.
Question: The man made an effort to impress his boss with the presentation to the client, because the _ was thoroughly researched.
Answer: presentation.
Output:
| [
"The man made an effort to impress his boss with the presentation to the client, because the _ was hard to please."
] | task034-bfef21db05694b84a7e6bbc511b4448e |
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: computer.
Question: The computer couldn't connect to the iphone because the _ is too old to be able to.
Answer: computer.
Output:
| [
"The computer couldn't connect to the iphone because the _ is too new to be able to."
] | task034-2dd9ecdf8aea4cc499c7b583fc52fa33 |
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: camera.
Question: One reason I switched from SLR cameras to mirror-less cameras because the _ are heavy.
Answer: SLR cameras.
Output:
| [
"One reason I switched from SLR cameras to mirror-less cameras because the _ are light."
] | task034-dc5984384d3d484d8dd178348eb6a011 |
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: Tommy could not put the book into the mailing box because the _ was too thick.
Answer: book.
Output:
| [
"Tommy could not put the book into the mailing box because the _ was too thin."
] | task034-98b7762578c9420fa93c08be5e22d538 |
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: purse.
Question: Sue tried place her old hair brush into her purse, but it would not fit because the _ was too long.
Answer: brush.
Output:
| [
"Sue tried place her old hair brush into her purse, but it would not fit because the _ was too short."
] | task034-3e80640622bb4f0fb4f15e9b6eaada9f |
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: pin.
Question: Mary needed to fix their hair by putting in a bobby pin, so now the _ is secured.
Answer: pin.
Output:
| [
"Mary needed to fix their hair by putting in a bobby pin, so now the _ is pinned."
] | task034-bf9ecf511da14797a1b4a148ffdb58b9 |
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: speech.
Question: The president had their speech at the auditorium rather than the park because the _ was bigger.
Answer: auditorium.
Output:
| [
"The president had their speech at the auditorium rather than the park because the _ was smaller."
] | task034-fa4fda2f642e48bb8f31c431ed0d1eea |
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: job.
Question: Sarah was unable to find the job opening that they wanted at their school because the _ is small.
Answer: school.
Output:
| [
"Sarah was unable to find the job opening that they wanted at their school because the _ is full."
] | task034-f83786c0c84746d5b1bf53ca116d7ab9 |
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: The fasting was a lot worse than the diet, because with the _ you couldn't eat at times.
Answer: fasting.
Output:
| [
"The fasting was a lot worse than the diet, because with the _ you could always eat at times."
] | task034-84be7b9d0dc6431abf8abe52fc2d2f04 |
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: strengthen.
Question: The weights and the squats would both strengthen her muscles, but the _ could be felt more the next day.
Answer: squats.
Output:
| [
"The weights and the squats would both strengthen her muscles, but the _ could be felt less the next day."
] | task034-a368943e26c149cfb295df78988f2d90 |
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: mirror.
Question: The man didn't want to get his rearview mirror replaced before the windshield, since the _ was dirtier.
Answer: windshield.
Output:
| [
"The man didn't want to get his rearview mirror replaced before the windshield, since the _ was cleaner."
] | task034-941b4e4356d5496a8b824d6e8698fe37 |
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: dirt.
Question: The plants grew better in the dirt near the tree than the dirt near the house because there was salt in the dirt near the _ .
Answer: house.
Output:
| [
"The plants grew better in the dirt near the tree than the dirt near the house because there was manure in the dirt near the _ ."
] | task034-5fd519f937d24291b78b7874777a7d90 |
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: bowl.
Question: The food keeps falling out when the dog was eating from the bowl. The _ is big.
Answer: food.
Output:
| [
"The food keeps falling out when the dog was eating from the bowl. The _ is small."
] | task034-0a7ebbd002904d1da4848021311dd9fa |
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: blade.
Question: He decided to forge a blade instead of buy a knife, because the _ would be a lot stronger and personal.
Answer: blade.
Output:
| [
"He decided to forge a blade instead of buy a knife, because the _ would be a lot weaker and less personal."
] | task034-19239f47b72b417bab3dbf1728ffa4bb |
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: sleeping bag.
Question: The man had no problem fitting the sleeping bag in the trunk because the _ was compact.
Answer: sleeping bag.
Output:
| [
"The man had no problem fitting the sleeping bag in the trunk because the _ was roomy."
] | task034-5325d7d242124e14bc2c21e4801e11fb |
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: orange.
Question: James could hardly squeezed the orange with the amount of force applied. The _ is hard.
Answer: orange.
Output:
| [
"James could hardly squeezed the orange with the amount of force applied. The _ is weak."
] | task034-b5be88b27b524ca9b5702a3e8ccf3883 |
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: swelling.
Question: Tim put ice and then heat on the swelling since _ would contract the area.
Answer: ice.
Output:
| [
"Tim put ice and then heat on the swelling since _ would loosen the area."
] | task034-f2f321ba042b4753a46b9d5986fd71cf |
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: method.
Question: I was unsure the method would work for the difficult task but it did because the _ was simple.
Answer: method.
Output:
| [
"I was unsure the method would work for the difficult task but it did because the _ was doable."
] | task034-7535210d0b2e4b30b835c3bab85cb740 |
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: rose.
Question: The rose I got was beautiful so I put it in the window rather than on the table, because others could see it from the _ more clearly.
Answer: window.
Output:
| [
"The rose I got was beautiful so I put it in the window rather than on the table, because others couldn't see it from the _ very clearly."
] | task034-7879baacd325432f84a77d00f1229322 |
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: pasta.
Question: Nora was boiling some pasta in a pot of water for dinner and made a mess. The _ she put in it was too high.
Answer: water.
Output:
| [
"Nora was boiling some pasta in a pot of water for dinner and made a mess. The _ she put it in was too low."
] | task034-c97f58626f65417d91211aa869eb47a8 |
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: innocent.
Question: After the accident, the innocent baby needed another bib but not clothes as the _ was torn.
Answer: bib.
Output:
| [
"After the accident, the innocent baby needed another bib but not clothes as the _ was sewn."
] | task034-7ccd75bd4d844abc9c5a5d51ee5eef71 |
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: egg.
Question: Ryan usually ate an egg for breakfast, but he didn't this morning because the _ was rotten.
Answer: egg.
Output:
| [
"Ryan usually ate an egg for breakfast, but he didn't this morning because the _ was cold."
] | task034-2b33f4af238e4b039a21346ce633336d |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.