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: doll.
Question: Cassie hates the doll I got her because I misheard whether she wanted a infant or a Barbie. The _ doll is for older kids.
Answer: barbie.
Output:
| [
"Cassie hates the doll I got her because I misheard whether she wanted a infant or a Barbie. The _ doll is for younger kids."
] | task034-507e2a7b7d2e426c9e198425edf4d20d |
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: wood.
Question: The axe handle can be made of wood or metal, the _ tends to crack over time.
Answer: wood.
Output:
| [
"The axe handle can be made of wood or metal, the _ tends to rustover time."
] | task034-9c64c6eb39a743a380c94fb8d5824ad8 |
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: eyebrows.
Question: Jan wanted to color her eyebrows in the bathroom light but the _ is too bright.
Answer: bathroom.
Output:
| [
"Jan wanted to color her eyebrows in the bathroom light but the _ is too dim."
] | task034-09f75352cf5943eabee75a962d075910 |
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: music.
Question: I love all sorts of music but rock and classical are my favorites. The _ music is best when you want to have a good time.
Answer: rock.
Output:
| [
"I love all sorts of music but rock and classical are my favorites. The _ music is best when you want to relax."
] | task034-312a6e27aee94f44897565820266519a |
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: waltz.
Question: We wanted to do the waltz to the rock song but we couldn't because the _ was too fast.
Answer: rock song.
Output:
| [
"We wanted to do the waltz to the rock song but we couldn't because the _ was too slow."
] | task034-6ca369e71e4145a683ee04d92bf54145 |
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: avocado.
Question: Jesse bought some unripe avocados and waited for them to get ripe. He waited long enough and the _ were ready.
Answer: ripe avocados.
Output:
| [
"Jesse bought some unripe avocados and waited for them to get ripe. He didn't wait long enough and the _ weren't ready."
] | task034-f3926c8f346e46bc87651c66bb27bd7f |
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: contest.
Question: The essay contest took no time compared to the exam the students had to take. The _ was easy.
Answer: essay.
Output:
| [
"The essay contest took no time compared to the exam the students had to take. The _ was hard."
] | task034-babcc122e8c148f49737aea3abc9318c |
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: draw.
Question: The boy liked to draw tigers, but not bears because _ are much more fascinating to the boy.
Answer: tigers.
Output:
| [
"The boy liked to draw tigers, but not bears because _ are much more dull to the boy."
] | task034-e7a1ee7ada994529b246e1902fa5f7de |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: store.
Question: Daniel went to the store late so the sale on mowers ended, so he found some tires instead. While his needs were disclosed, the _ were available.
Answer: tires.
Output:
| [
"Daniel went to the store late so the sale on mowers ended, so he found some tires instead. While his needs were disclosed, the _ were unobtainable."
] | task034-f9a1453440094aa482e7796d6634b2ac |
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: fire.
Question: Mary's apartment had a door in front and a kitchen window in back. When a fire broke out in the kitchen, Mary escaped through the _ .
Answer: door.
Output:
| [
"Mary's apartment had a door in front and a kitchen window in back. When a fire broke out in the entryway, Mary escaped through the _ ."
] | task034-0947a2b3faf34a9fa292f0a965057d51 |
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: photos.
Question: When she examined the newly developed photos, the trees looked odd because the _ were short.
Answer: trees.
Output:
| [
"When she examined the newly developed photos, the trees looked odd because the _ were bright."
] | task034-69a31bc2d5314e20bca4d15781b2e799 |
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: file.
Question: The file could have got burnt during the fire outbreak but thanks to the cabinet in which it was kept. The _ is inflammable.
Answer: file.
Output:
| [
"The file could have got burnt during the fire outbreak but thanks to the cabinet in which it was kept. The _ is flammable."
] | task034-a5fd87cea5aa4c9db52babf56f6ce186 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: color.
Question: I like to shoot photography in RAW format compared to JPEG format because the _ files have narrow dynamic color ranges.
Answer: JPEG.
Output:
| [
"I like to shoot photography in RAW format compared to JPEG format because the _ files have wide dynamic color ranges."
] | task034-1bbe5e793abd4b10972bff0dc5327412 |
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: doors.
Question: I was able to easily fit my tables through the the double doors. The _ were quite wide.
Answer: doors.
Output:
| [
"I was able to easily fit my tables through the the double doors. The _ were quite narrow."
] | task034-5c52b3bc27b24ce98d39bd93264ad1d8 |
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: diapers.
Question: Having outgrown diapers, the child switched to shorts but not pants, as the _ fit perfectly.
Answer: shorts.
Output:
| [
"Having outgrown diapers, the child switched to shorts but not pants, as the _ fit poorly."
] | task034-af42b087859147d9b5cdb9c11c713784 |
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: pancakes.
Question: Bob enjoyed his pancakes with butter and bacon, but he was upset when the _ arrived too hard.
Answer: butter.
Output:
| [
"Bob enjoyed his pancakes with butter and bacon, but he was upset when the _ arrived too soft."
] | task034-e918ff51694640a097b2ea9999ad17ae |
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: stock.
Question: James wanted to buy the stock for the company when given advice but the _ is too unstable.
Answer: stock.
Output:
| [
"James wanted to buy the stock for the company when given advice because the _ is stable."
] | task034-022ca9fddc554b85b1fb1644692779ba |
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: reach.
Question: John had to reach both the neighbors home and the ocean today, but thankfully the _ was close.
Answer: neighbors home.
Output:
| [
"John had to reach both the neighbors home and the ocean today, but regrettable the _ was close."
] | task034-e94a75ba461c4ea2a300cf31f622e117 |
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: leather.
Question: Laura's distressed leather shoes matched her belt but not her purse, as the _ was patent leather.
Answer: purse.
Output:
| [
"Laura's distressed leather shoes matched her belt but not her purse, as the _ was distressed leather."
] | task034-e62c99ed23904c9591d51133bf030576 |
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: John wanted to to lift more weight with his muscles but the _ are too big.
Answer: weight.
Output:
| [
"John wanted to to lift more weight with his muscles but the _ are too slim."
] | task034-ae6d016356d8486bb55e4e5fce910dea |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: bike.
Question: Jake replaced his metal bike for a carbon one because the _ was too heavy.
Answer: metal.
Output:
| [
"Jake replaced his metal bike for a carbon one because the _ was very lightweight."
] | task034-ea5bd3db63704f958ed6290fd576c2ee |
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: grades.
Question: I need to raise my grades. I could stop playing my favorite game and start studying. The _ is distracting.
Answer: game.
Output:
| [
"I need to raise my grades. I could stop playing my favorite game and start studying. The _ is beneficial."
] | task034-a6581cbb05b04036b0aff2f79a616109 |
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: pregnant.
Question: Vera craved steamed asparagus while she was pregnant but ditched the pepperoni pizza. Therefore the _ was wasted.
Answer: pizza.
Output:
| [
"Vera craved steamed asparagus while she was pregnant but ditched the pepperoni pizza. Therefore the _ was devoured."
] | task034-f6779b598e6c4005835c31545b7d758a |
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: pregnancy.
Question: After finding about the pregnancy I started hoarding diapers and clothes. Mostly the _ because this will be 2nd child.
Answer: diapers.
Output:
| [
"After finding about the pregnancy I started hoarding diapers and clothes. Mostly the _ because the other will be easy."
] | task034-49af750fcb6843d68e61cf5b388056f0 |
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: loner.
Question: When deciding if to be a loner, one must decide if they want many events or plenty of free time. The _ angle helps with productivity.
Answer: free time.
Output:
| [
"When deciding if to be a loner, one must decide if they want many social events or plenty of free time. The _ angle helps with sociability."
] | task034-0b7e22b80fc14b75b7d49ca5d4ca8c72 |
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: guitar.
Question: The band member tried to put the guitar in the case but it wouldn't fit because the _ was too big.
Answer: guitar.
Output:
| [
"The band member tried to put the guitar in the case but it wouldn't fit because the _ was too small."
] | task034-0f7d8f2f28c6462f95bbc3d96f42a334 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: work.
Question: My sister hated her work a lot more than I hated my job, because the _ was difficult.
Answer: work.
Output:
| [
"My sister hated her work a lot more than I hated my job, because the _ was easy."
] | task034-17916dc9e9cc44d48c4ab726a06f2c5b |
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: sound.
Question: The sound from the train was louder than the sound from the car because the _ had a bigger engine.
Answer: train.
Output:
| [
"The sound from the train was louder than the sound from the car because the _ had a smaller engine."
] | task034-cc0b986b0deb4a64aee10293b547a0c4 |
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: video game.
Question: Tony bought a video game at the store by taking a bus ride but the _ was too easy.
Answer: game.
Output:
| [
"Tony bought a video game at the store by taking a bus ride but the _ was too difficult."
] | task034-fcc4c6c706be49f3b0361b8a58ebba79 |
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: convince.
Question: She had to convince them to go to the concert instead of the park, but the _ was full.
Answer: concert.
Output:
| [
"She had to convince them to go to the concert instead of the park, but the _ was free."
] | task034-bb84f821a3e94c208f327ec9421d4545 |
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: Photo.
Question: Liz was eager to take a photo with her new camera and flash but the _ was too dark.
Answer: photo.
Output:
| [
"Liz was eager to take a photo with her new camera and flash but the _ was too bright."
] | task034-76c9dccc733449a1b2ce83f2468448f5 |
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: disorder.
Question: They were hoping that it was just an eating disorder instead of a disease, because the _ would be deadly.
Answer: disease.
Output:
| [
"They were hoping that it was just an eating disorder instead of a disease, because the _ would be survivable."
] | task034-b968cbd106554a11930cf77a733778d1 |
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: balance.
Question: It is easier for the young to keep their balance than the very venerable as the _ tend to have fewer health problems.
Answer: young.
Output:
| [
"It is easier for the young to keep their balance than the very venerable as the _ tend to have more health problems."
] | task034-8191c733d9ed466ea9dd1d472b1c6edd |
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: ball.
Question: Even though it traveled on target, the ball could not fit in the hole because the _ was too big.
Answer: ball.
Output:
| [
"Even though it traveled on target, the ball could not fit in the hole because the _ was too small."
] | task034-804a3c4977f34bbe9a3cac1204c585b2 |
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: cat.
Question: Cincy wanted to show affection to her cat, so she bought it a toy instead of food, because the cat finds the _ to be bad.
Answer: food.
Output:
| [
"Cincy wanted to show affection to her cat, so she bought it a toy instead of food, because the cat finds the _ to be good."
] | task034-1226b75c5e724a72aba8fc865a190fa3 |
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: Shampoo.
Question: Harold reached for the shampoo before the conditioner because the _ was less accessible to him.
Answer: conditioner.
Output:
| [
"Harold reached for the shampoo before the conditioner because the _ was more accessible to him."
] | task034-b8fd9a1b03f844f7bff8427d2783fed2 |
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: health.
Question: Gin opted to make changes in his health and cut meat out of his diet because the _ was in a dire state.
Answer: health.
Output:
| [
"Gin opted to make changes in his health and cut meat out of his diet because the _ was in an improvable state."
] | task034-37545762ef8e43e6bb3634903f8e49a3 |
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: perfume.
Question: Phillipe wanted to use a perfume but the nozzle was blocked because of dirt. Luckily he could clean the _ out of the it.
Answer: dirt.
Output:
| [
"Phillipe wanted to use a perfume but the nozzle was blocked because of dirt. Luckily he could remove _ out of the bottle."
] | task034-0fd26bcaf143496380cb73ee01c51700 |
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: revolver.
Question: John went to the shooting range he tried the revolver and the rifle. The _ hurt his shoulder.
Answer: rifle.
Output:
| [
"John went to the shooting range he tried the revolver and the rifle. The _ hurt his hand."
] | task034-bdb9a5dec9cd49b1b8e569f46dddabfd |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: car.
Question: The fuel inside the car is not enough for the journey. I never know the _ would be this long.
Answer: journey.
Output:
| [
"The fuel inside the car is not enough for the journey. I never know the _ would be this small."
] | task034-0f319920481441fc8b99443efddb073b |
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: wall.
Question: It is so much easier to remove a wall made of drywall than plaster because the _ does not crumble.
Answer: drywall.
Output:
| [
"It is so much easier to remove a wall made of drywall than plaster because the _ crumbles."
] | task034-9481e4408e7c498cbc083fe4bf1a425d |
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: romantic.
Question: Tim bought his wife flowers instead of candy because he thought the _ was less romantic.
Answer: candy.
Output:
| [
"Tim bought his wife flowers instead of candy because he thought the _ was more romantic."
] | task034-a949f45f347d4bccb39b7aa8fb1680b1 |
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: furnace.
Question: Both the furnace and the air conditioner were turned off, so Janet turned on the _ because it was hot.
Answer: air conditioner.
Output:
| [
"Both the furnace and the air conditioner were turned off, so Janet turned on the _ because it was cold."
] | task034-82aab1b9922f4d16b6761cf2a4d63ed8 |
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: savings.
Question: I prefer to invest my savings in Mutual funds and not in Tax-free savings. The _ has higher interest rate.
Answer: Mutual funds.
Output:
| [
"I prefer to invest my savings in Mutual funds and not in Tax-free savings. The _ has lower interest rate."
] | task034-58a5394442ba4192b4e4b2cd5c136e11 |
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: Jason really wanted to get the job, but he felt like he flunked the interview. The manager called and said he could start the _ next week.
Answer: job.
Output:
| [
"Jason really wanted to get the job, but he felt like he flunked the interview. The manager called and said he could try the _ again next week."
] | task034-602a8d1903604f12b7f30cb5171cdc66 |
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: athlete.
Question: The athlete was at the gym more than he was at the beach because the _ was where he trained.
Answer: athlete.
Output:
| [
"The athlete was at the gym more than he was at the beach although the _ was where he trained."
] | task034-26ea7d1e31f84d54985c01d6a3ff610c |
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 shake was much more effective for building muscle than the sandwich, since the _ had sufficient nutrients.
Answer: shake.
Output:
| [
"The shake was much more effective for building muscle than the sandwich, since the _ had inadequate nutrients."
] | task034-bec8fef1372649768c7a012fb329960c |
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: sports.
Question: Kim wanted to play sports but either her shoulders or her back prevented her, she could not bend over so she knew it was the _ .
Answer: back.
Output:
| [
"Kim wanted to play sports but either her shoulders or her back prevented her, she could not reach over so she knew it was the _ ."
] | task034-9db4389887c243f59984360f12634d28 |
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: supplements.
Question: Steroids are a much worse choice to use than supplements, because the _ have many side effects.
Answer: steroids.
Output:
| [
"Steroids are a much worse choice to use than supplements, because the _ have few side effects."
] | task034-96f2d88f10c242e4a701d104d80abdce |
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: cat.
Question: Sheldon bought the food for his cat instead of the vodka, since buying the _ was the more irresponsible choice to make.
Answer: vodka.
Output:
| [
"Sheldon bought the food for his cat instead of the vodka, since buying the _ was the more responsible choice to make."
] | task034-5f56d9d17aba408e8386aebac619705d |
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: care.
Question: My houseplants and vegetable garden needed tender loving care. The _ were dying from too much shade.
Answer: houseplants.
Output:
| [
"My houseplants and vegetable garden needed tender loving care. The _ were dying from too much sun."
] | task034-f60e3d8ba6bc4b528c44d47d44b9deb3 |
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: pack.
Question: The family packed clothes in the car for the trip, though the _ didn't fit them all.
Answer: car.
Output:
| [
"The family packed clothes in the car for the trip, though the _ didn't fit inside."
] | task034-351227c8d0a5423abeaee0b7c8db8946 |
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: illness.
Question: A respiratory illness is potentially dangerous in an unhealthy person but not so much in a healthy person as the _ person has the worse immune system.
Answer: unhealthy.
Output:
| [
"A respiratory illness is potentially dangerous in an unhealthy person but not so much in a healthy person as the _ person has the better immune system."
] | task034-14ad022ad5be4ed1b5027702540c9558 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: bedroom.
Question: The cat peed in the bedroom a lot while the dog never did. _ was badly behaved.
Answer: cat.
Output:
| [
"The cat peed in the bedroom a lot while the dog never did. _ was well behaved."
] | task034-425f04e9565c48e49b82545328bfa59c |
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: soda.
Question: The price of soda was increasing at the store but going down at the market, since the _ had a shortage.
Answer: store.
Output:
| [
"The price of soda was increasing at the store but going down at the market, since the _ had a surplus."
] | task034-8ec4efc77daf43a68b8a48aff8dc4469 |
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: youtube.
Question: Jenny wanted more hits on her youtube channel so she posted more content but the _ was too long.
Answer: content.
Output:
| [
"Jenny wanted more hits on her youtube channel so she posted more content but the _ was too boring."
] | task034-71a054af668a44e7941e37f0bf26f128 |
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: cage.
Question: Jill bought a rock as decoration for the cage, but it didn't work because the _ was too small.
Answer: cage.
Output:
| [
"Jill bought a rock as decoration for the cage, but it didn't work because the _ was too big."
] | task034-998a9503e05447b6acdd913b694a316e |
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: snow.
Question: I preferred snow to the cold rain because the _ was more pretty and did not soak me to the bone.
Answer: snow.
Output:
| [
"I preferred snow to the cold rain because the _ was not pretty and did soak me to the bone."
] | task034-7779f40d2ccb4220846811ec61239292 |
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: chickenpox.
Question: The boy had contracted the chickenpox virus that was going around the classroom, because the _ was very infectious.
Answer: virus.
Output:
| [
"The boy had contracted the chickenpox virus that was going around the classroom, because the _ was very small."
] | task034-364b405fe25e44f6bb1db81fffcf8bc5 |
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: Lights.
Question: Adam put the lights up on the trees for Christmas because the _ would make them more decorated.
Answer: lights.
Output:
| [
"Adam put the lights up on the trees for Christmas because the _ would needed to be decorated."
] | task034-c3e187e0335e4e8488389b773ec55b0c |
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: bleeding.
Question: The heavy bleeding coming out of the nose was entering the mouth because the _ was expelling so much blood.
Answer: nose.
Output:
| [
"The heavy bleeding coming out of the nose was entering the mouth because the _ was absorbing so much blood."
] | task034-87ab115064f848398f7f4b303fac0357 |
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: James squeezed the water out of the towel to the floor and it get the _ dry.
Answer: towel.
Output:
| [
"James squeezed the water out of the towel to the floor and it get the _ wet."
] | task034-4a26387eca174787b0ba083cb3610977 |
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: journal.
Question: Erin wanted to keep her journal in her purse for easy access but it didn't fit. The _ was too wide.
Answer: journal.
Output:
| [
"Erin wanted to keep her journal in her purse for easy access but it didn't fit. The _ was too thin."
] | task034-784c5c3233724afe9216854122985bae |
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: window.
Question: John opened the window out of the curtain and the light ray comes in. The _ is transparent.
Answer: window.
Output:
| [
"John opened the window out of the curtain and the light ray comes in. The _ is opaque."
] | task034-3cefea2eae35401082c37d71d3f455ce |
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: camping.
Question: Kai wanted to transport the tent by his car when he went to the camping, it worked great as the _ was big enough.
Answer: car.
Output:
| [
"Kai wanted to transport the tent by his car when he went to the camping, it worked great as the _ was small when folded."
] | task034-20cdcf31fe864beaabed19fa60f924f0 |
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: running.
Question: The water was running out of the hose and into the bucket, because the _ was receiving liquid.
Answer: bucket.
Output:
| [
"The water was running out of the hose and into the bucket, because the _ was providing liquid."
] | task034-f6b34fcc3f3740759c28d2c342273ce2 |
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: comfortable.
Question: The boy felt comfortable in his blanket while staying home from school. The _ cozy and hard.
Answer: home.
Output:
| [
"The boy felt comfortable in his blanket while staying home from school. The _ small and soft."
] | task034-752f451025f4403bb06d2c00cd8a3543 |
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: nails.
Question: Scott used nails from his toolbox to put the shingles on the roof until the _ was finished.
Answer: roof.
Output:
| [
"Scott used nails from his toolbox to put the shingles on the roof until the _ was empty."
] | task034-ed6d81c1b02a4a93acac814102e12bfb |
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: lavender.
Question: The home was adorned with the scent of lavender by candles and sprays. The _ smelled natural.
Answer: candles.
Output:
| [
"The home was adorned with the scent of lavender by candles and sprays. The _ smelled chemical."
] | task034-cb1fafcb85bc432c9bf0461532491da9 |
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: foot.
Question: The arrow missed the target by only a foot, as the _ was aimed poorly.
Answer: arrow.
Output:
| [
"The arrow missed the target by only a foot, as the _ was positioned poorly."
] | task034-3c23774152e144489508adfb0cdb680a |
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: journal.
Question: They thought the critic was dramatic after writing a big article about the journal because the _ was so short.
Answer: journal.
Output:
| [
"They thought the critic was dramatic after writing a big article about the journal because the _ was so long."
] | task034-7296a3bc3b1c4b5393d824eb6e418886 |
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: boil.
Question: The small scratch turned into a festering boil that required antibiotics, because the _ became infected.
Answer: boil.
Output:
| [
"The small scratch turned into a festering boil that required antibiotics, because the _ treated the infection."
] | task034-20d4868b49dd46a49d439eb157432fca |
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: study.
Question: Layla had a test coming up but was unable to study their study guide enough because the _ was soon.
Answer: test.
Output:
| [
"Layla had a test coming up but was unable to study their study guide enough because the _ was long."
] | task034-d7ada40449af4634ad2e0afa8b55b213 |
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: cats.
Question: The cats shed more than the dogs because the _ had more hair on them.
Answer: cats.
Output:
| [
"The cats shed more than the dogs because the _ had less hair on them."
] | task034-9719afbdb7d6474a8ebace1dc8fce607 |
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: diamonds.
Question: Janine switched from selling cooper and silver necklaces to selling diamonds. The _ are more expensive.
Answer: diamonds.
Output:
| [
"Janine switched from selling cooper and silver necklaces to selling diamonds. The _ are cheaper.."
] | task034-15a1cb5c66d04b7eb261624f1bcbf5ed |
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: paint.
Question: James probably got stained with the paint from the gallon and not the wall because the _ is dry.
Answer: wall.
Output:
| [
"James probably got stained with the paint from the gallon and not the wall because the _ is wet."
] | task034-678737d4854645efa2a71b2056c2e197 |
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: dating.
Question: Tim wanted to start dating a girl so he decided to wait for her at the bust stop that afternoon but the _ was too short.
Answer: afternoon.
Output:
| [
"Tim wanted to start dating a girl so he decided to wait for her at the bust stop that afternoon but the _ was too long."
] | task034-5a3009fcfe474a158ca5d334a3658eb4 |
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: split.
Question: James crushed the pod under the force of his finger while trying to split it into two because the _ is too soft.
Answer: pod.
Output:
| [
"James crushed the pod under the force of his finger while trying to split it into two because the _ is too hard."
] | task034-f8ecfffe357f40d88f7bb191d13cce7f |
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: film.
Question: I need either an action or romance film to watch with my boyfriend. He's actually into that sappy stuff so maybe the _ film.
Answer: romance.
Output:
| [
"I need either an action or romance film to watch with my boyfriend. He's actually into that explosion stuff so maybe the _ film."
] | task034-b7d17f8db58f4ea8ad506ac03c6a2fb7 |
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: tape.
Question: Jen taped her pointer finger to her middle finger, because the _ was broken and she wanted it to heal.
Answer: pointer finger.
Output:
| [
"Jen taped her pointer finger to her middle finger, because the _ was supporting the finger she wanted to heal."
] | task034-783b91b343b34c89b055117a00944be2 |
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: skin treatment.
Question: The skin treatment made her skin look better than the face mask did. The _ must work efficiently.
Answer: skin treatment.
Output:
| [
"The skin treatment made her skin look better than the face mask did. The _ must work poorly."
] | task034-eaa93cb103474a769e30ac591611847c |
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: She tried to write her essay on the computer and with the ink pen but the _ was broken.
Answer: computer.
Output:
| [
"She tried to write her essay on the computer and with the ink pen but the _ was empty."
] | task034-b7acfdf27c83497eae0177b92a12f175 |
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: catfish.
Question: Rob wanted catfish for dinner, but he couldn't decide on grilled or fried. He knew that the _ was healthier.
Answer: grilled.
Output:
| [
"Rob wanted catfish for dinner, but he couldn't decide on grilled or fried. He knew that the _ was fattier."
] | task034-f88b0ddf94eb4945802b602f40e72ff7 |
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: skin.
Question: It's important to use sunscreen on your skin on your face and shoulders. The _ will burn if you aren't wearing a hat.
Answer: face.
Output:
| [
"It's important to use sunscreen on your skin on your face and shoulders. The _ will burn if you aren't wearing a shirt."
] | task034-889f4d59d67a47b68a30a9ca33ebe87e |
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: basement.
Question: The basement was dark and dirty but the upstairs was renovated. The _ was scary.
Answer: basement.
Output:
| [
"The basement was dark and dirty but the upstairs was renovated. The _ was beautiful."
] | task034-0edf529e9684467eb01f54cf8e2ef1e7 |
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: daycare.
Question: The mother brought more diapers than bottles to the daycare center, because more of the _ were needed.
Answer: diapers.
Output:
| [
"The mother brought more diapers than bottles to the daycare center, because less of the _ were needed."
] | task034-2c8e8d8b5f544084ab03c5f575e535d4 |
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: decorate.
Question: She wanted to decorate her car but leave the house plain , since she needed the _ to impress people.
Answer: car.
Output:
| [
"She wanted to decorate her car but leave the house plain , since she didn't want the _ to impress people."
] | task034-9552f4ead2e74a43b9416175a9aec5d5 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: incense.
Question: The smoke of the incense filled the whole room and started to choke people because the _ is too much.
Answer: smoke.
Output:
| [
"The smoke of the incense filled the whole room and started to choke people because the _ is too small."
] | task034-6dab228ca87141d08112ce1e085175f8 |
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: wall.
Question: The man could not jump over the wall, but easily hopped over the fence, because the _ was taller.
Answer: wall.
Output:
| [
"The man could not jump over the wall, but easily hopped over the fence, because the _ was shorter."
] | task034-83913a9c4979432fb818926d5fb183f0 |
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: Ring.
Question: Susan had to take the loose ring off her skinny finger and give it to the jeweler to be sized because the _ was too big.
Answer: ring.
Output:
| [
"Susan had to take the loose ring off her skinny finger and give it to the jeweler to be sized because the _ was too small."
] | task034-a1422fee673343b59af334faf8797090 |
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: wash.
Question: James washed his hand inside the big bowl after eating and the _ gets dirty.
Answer: bowl.
Output:
| [
"James washed his hand inside the big bowl after eating and the _ gets clean."
] | task034-c68dcace6a5c4e1b9a5dd91fdd355c6d |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: morning.
Question: Jane did not want to get out of bed this morning because it was a rainy dark _ .
Answer: morning.
Output:
| [
"Jane did not want to get out of bed this morning because it was a warm comfy _ ."
] | task034-ab26e976307c4e0d995f5d3e7664ad96 |
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: ballet.
Question: They went to the ballet to see the performance live on stage, and the _ was cancelled due to an illness with the lead performer.
Answer: ballet.
Output:
| [
"They went to the ballet to see the performance live on stage, and the _ will continue in spite of an illness with the lead performer."
] | task034-d2025f7367054f9296a395f09ccb92f3 |
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: loves.
Question: Chrissy hates champagne, but loves her some wine. The _ industry is a scam to her.
Answer: champagne.
Output:
| [
"Chrissy hates champagne, but loves her some wine. The _ industry is a delight to her."
] | task034-6011306d5d0140d788d3e8184df78e4d |
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: soft hands.
Question: Hans wanted to have soft hands so she bought some expensive lotion with aloe but the _ was too much.
Answer: lotion.
Output:
| [
"Hans wanted to have soft hands so she bought some expensive lotion with aloe but the _ was too little."
] | task034-c940ed628afa4434b391780293bcc994 |
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: dance.
Question: The dance that was being taught to the entire dance class instead of the individuals was tough, so the _ were tired.
Answer: class.
Output:
| [
"The dance that was being taught to the entire dance class instead of the individuals was tough, so the _ were lively."
] | task034-5f9e59efd2f848b8b5fa2ccad162b234 |
Definition: In this task, you're given a fill-in-the-blank question that contains two object names. Additionally, you're given one answer which is one of the objects present in the question. In this task, you need to minimally change the given question so that the answer flips to another object in the question. Remember the question has to be about both objects which are related but different; for example, "trophy" and "suitcase". This task typically involves replacing one word i.e., the 'trigger word' with its antonym (e.g., changing from "small" to "big"). You should not change any content in the given question beyond a word or two, i.e., the trigger word/phrase. The expected answer to your question must not be associated with just the trigger word; instead, it should depend on the context present in the question. The expected answer should not be ambiguous. For your question, there should be an agreed upon answer to fill in the blank. Your generations should NOT contain potentially explicit, offensive, or adult content. In your generated question, retain the object names from the input question. Avoid repeating the same style or phrase in generating your modified question e.g. this task can be always solved using a simple negation i.e. by adding not, never, etc. Instead, try to increase the word diversity. Your question must contain at least 15 and at most 30 words. Your question must have at least 70% overlapping words with the given question. You must utilize the given context word while writing the question. Your question must contain only one blank. The two objects should be used ONCE in your question. Here is a list of attributes and associated contrastive words that may help write contrastive trigger words. Note that contrastive trigger words help in flipping the label.
| Attribute | triggerword | contrastive triggerword |
| age | old | new |
| altitude | low | high |
| area | small | vast |
| brightness | dark | light |
| clarity | obscure | clear |
| cleanness | dirty | clean |
| complexity | simple | complex |
| cost | cheap | expensive |
| density | sparse | dense |
| depth | shallow | deep |
| distance | near | far |
| electric conductivity | low | high |
| flexibility | rigid | flexible |
| granularity | fine | coarse |
| hardness | soft | hard |
| length | short | long |
| magnitude | small | large |
| mass | small | large |
| odor | weak | strong |
| pressure | low | high |
| resistance | low | high |
| shape | round | sharp |
| shape | flat | spiky |
| size | small | large |
| sound | quiet | loud |
| sound pitch | low | high |
| speed | slow | fast |
| stability | unstable | stable |
| strength | weak | strong |
| temperature | low | high |
| texture | smooth | rough |
| thermal conductivity | low | high |
| thickness | thin | thick |
| volume | small | large |
| weight | light | heavy |
| width | narrow | wide |
| location | in | out |
| location | up | down |
| location | above | below |
| location | on | off |
| location | to | from |
Positive Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because _ is too small.
Positive Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ looks great on her.
Negative Example 1 -
Input: Context word: fit.
Question: The trophy doesn't fit into the brown suitcase because _ is too large.
Answer: trophy.
Output: The trophy doesn't fit into the brown suitcase because there were a lot of other eseential items and jewlleries to pack and _ is too large.
Negative Example 2 -
Input: Context word: trade.
Question: Grace was happy to trade me her sweater for my jacket. She thinks _ looks dowdy on her.
Answer: sweater.
Output: Grace was happy to trade me her sweater for my jacket. She thinks _ is colored purple.
Now complete the following example -
Input: Context Word: friends.
Question: Steve was able to invite more friends to his house than he could when he lived in an apartment because the _ was smaller.
Answer: apartment.
Output:
| [
"Steve was able to invite more friends to his house than he could when he lived in an apartment because the _ was larger."
] | task034-7e64ad5c17ea4d0e9f90beeeb8f31e36 |
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: photo.
Question: John replaced the photo of his wedding with the painting because the _ is old.
Answer: photo.
Output:
| [
"John replaced the photo of his wedding with the painting because the _ is new."
] | task034-1e5cd414b8a04489acda83d80fa04480 |
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: television.
Question: My radio gets a lot less use than my television, because the _ is boring.
Answer: radio.
Output:
| [
"My radio gets a lot less use than my television, because the _ is entertaining."
] | task034-4574344f19c24046b6d8bee16111a161 |
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: pregnant.
Question: The pregnant woman wore dresses more than she wore jeans because the _ were tight on her.
Answer: jeans.
Output:
| [
"The pregnant woman wore dresses more than she wore jeans because the _ were loose on her."
] | task034-f2f39f00fcc7424981c46d711071ffd9 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.