skyyyyks/winoground_test · Datasets at Hugging Face

messages listlengths 2 2	images images listlengths 2 2
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a young person kisses an old person\"\n2) \"an old person kisses a young person\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a young person kisses an old person\n<image1>: an old person kisses a young person", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the shorter person hugs the taller person\"\n2) \"the taller person hugs the shorter person\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the shorter person hugs the taller person\n<image1>: the taller person hugs the shorter person", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the unmasked wrestler hits the masked wrestler\"\n2) \"the masked wrestler hits the unmasked wrestler\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the unmasked wrestler hits the masked wrestler\n<image1>: the masked wrestler hits the unmasked wrestler", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"an animal watches a person\"\n2) \"a person watches an animal\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: an animal watches a person\n<image1>: a person watches an animal", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with earrings pays the person without earrings\"\n2) \"the person without earrings pays the person with earrings\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with earrings pays the person without earrings\n<image1>: the person without earrings pays the person with earrings", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a snake eats a bird\"\n2) \"a bird eats a snake\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a bird eats a snake\n<image1>: a snake eats a bird", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"another organism was harmed by a plant, and that plant broke the organism into pieces\"\n2) \"a plant was harmed by another organism, and that organism broke the plant into pieces\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: another organism was harmed by a plant, and that plant broke the organism into pieces\n<image1>: a plant was harmed by another organism, and that organism broke the plant into pieces", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"water is in a bottle\"\n2) \"a bottle is in water\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: water is in a bottle\n<image1>: a bottle is in water", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a tree smashed into a car\"\n2) \"a car smashed into a tree\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a car smashed into a tree\n<image1>: a tree smashed into a car", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"people fall on the leaves\"\n2) \"the leaves fall on people\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the leaves fall on people\n<image1>: people fall on the leaves", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the water rests below the sail\"\n2) \"the sail rests below the water\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the sail rests below the water\n<image1>: the water rests below the sail", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is someone below a table\"\n2) \"there is a table below someone\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is a table below someone\n<image1>: there is someone below a table", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the image shows a computer on top of books\"\n2) \"the image shows books on top of a computer\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the image shows a computer on top of books\n<image1>: the image shows books on top of a computer", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a plant with some caterpillars\"\n2) \"a caterpillar with some plants\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a caterpillar with some plants\n<image1>: a plant with some caterpillars", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is some grass in a mug\"\n2) \"there is a mug in some grass\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is some grass in a mug\n<image1>: there is a mug in some grass", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are more snowboarders than skiers\"\n2) \"there are more skiers than snowboarders\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are more skiers than snowboarders\n<image1>: there are more snowboarders than skiers", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are fewer spoons than forks\"\n2) \"there are fewer forks than spoons\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are fewer spoons than forks\n<image1>: there are fewer forks than spoons", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are more ladybugs than flowers\"\n2) \"there are more flowers than ladybugs\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are more ladybugs than flowers\n<image1>: there are more flowers than ladybugs", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is less milk than orange juice\"\n2) \"there is less orange juice than milk\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is less milk than orange juice\n<image1>: there is less orange juice than milk", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is more dirt than empty space in the jar\"\n2) \"there is more empty space than dirt in the jar\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is more empty space than dirt in the jar\n<image1>: there is more dirt than empty space in the jar", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is a bit less white chocolate than milk chocolate\"\n2) \"there is a bit less milk chocolate than white chocolate\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is a bit less milk chocolate than white chocolate\n<image1>: there is a bit less white chocolate than milk chocolate", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a person stands and a dog sits\"\n2) \"a person sits and a dog stands\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a person sits and a dog stands\n<image1>: a person stands and a dog sits", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"The person with more facial hair lies awake while the other sleeps\"\n2) \"The person with more facial hair sleeps while the other lies awake\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: The person with more facial hair lies awake while the other sleeps\n<image1>: The person with more facial hair sleeps while the other lies awake", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person in the middle is standing and some others are floating\"\n2) \"the person in the middle is floating and some others are standing\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person in the middle is standing and some others are floating\n<image1>: the person in the middle is floating and some others are standing", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person closer to the camera weightlifts and the person farther from the camera runs\"\n2) \"the person closer to the camera runs and the person farther from the camera weightlifts\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person closer to the camera weightlifts and the person farther from the camera runs\n<image1>: the person closer to the camera runs and the person farther from the camera weightlifts", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with the white collared shirt holds the plant while the other waters it\"\n2) \"the person with the white collared shirt waters the plant while the other holds it\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the white collared shirt waters the plant while the other holds it\n<image1>: the person with the white collared shirt holds the plant while the other waters it", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the taller person chops food and the shorter person eats food\"\n2) \"the taller person eats food and the shorter person chops food\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the taller person eats food and the shorter person chops food\n<image1>: the taller person chops food and the shorter person eats food", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with the ponytail packs stuff and other buys it\"\n2) \"the person with the ponytail buys stuff and other packs it\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the ponytail buys stuff and other packs it\n<image1>: the person with the ponytail packs stuff and other buys it", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a person walking in a field of red flowers with a yellow dress\"\n2) \"a person walking in a field of yellow flowers with a red dress\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a person walking in a field of yellow flowers with a red dress\n<image1>: a person walking in a field of red flowers with a yellow dress", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a white dog is on a brown couch\"\n2) \"a brown dog is on a white couch\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a white dog is on a brown couch\n<image1>: a brown dog is on a white couch", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a red bird is next to a blue berry\"\n2) \"a blue bird is next to a red berry\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a red bird is next to a blue berry\n<image1>: a blue bird is next to a red berry", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"The red book is above the yellow book and below the blue book\"\n2) \"The yellow book is above the blue book and below the red book\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: The yellow book is above the blue book and below the red book\n<image1>: The red book is above the yellow book and below the blue book", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"astronauts in blue suits with a red planet in the background\"\n2) \"astronauts in red suits with a blue planet in the background\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: astronauts in blue suits with a red planet in the background\n<image1>: astronauts in red suits with a blue planet in the background", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the square table has a round base\"\n2) \"the round table has a square base\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the square table has a round base\n<image1>: the round table has a square base", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"rectangular bushes are behind pointy bushes\"\n2) \"pointy bushes are behind rectangular bushes\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: rectangular bushes are behind pointy bushes\n<image1>: pointy bushes are behind rectangular bushes", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a rectangular sign is above a circular sign\"\n2) \"a circular sign is above a rectangular sign\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a rectangular sign is above a circular sign\n<image1>: a circular sign is above a rectangular sign", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a metal chess piece rests on wood objects\"\n2) \"a wood chess piece rests on metal objects\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a metal chess piece rests on wood objects\n<image1>: a wood chess piece rests on metal objects", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"soft shoes are on a smooth floor\"\n2) \"smooth shoes are on a soft floor\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: smooth shoes are on a soft floor\n<image1>: soft shoes are on a smooth floor", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the happy person is on the right and the sad person is on the left\"\n2) \"the sad person is on the right and the happy person is on the left\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the sad person is on the right and the happy person is on the left\n<image1>: the happy person is on the right and the sad person is on the left", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the hurt person is on the left and the helpful person is on the right\"\n2) \"the helpful person is on the left and the hurt person is on the right\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the hurt person is on the left and the helpful person is on the right\n<image1>: the helpful person is on the left and the hurt person is on the right", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the less content person has the healthier food and the more content person has the less healthy food\"\n2) \"the more content person has the healthier food and the less content person has the less healthy food\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the more content person has the healthier food and the less content person has the less healthy food\n<image1>: the less content person has the healthier food and the more content person has the less healthy food", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the orange lollipop is surprised and the red lollipop is sad\"\n2) \"the orange lollipop is sad and the red lollipop is surprised\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the orange lollipop is sad and the red lollipop is surprised\n<image1>: the orange lollipop is surprised and the red lollipop is sad", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the smaller person wears yellow and the larger person does not\"\n2) \"the larger person wears yellow and the smaller person does not\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the smaller person wears yellow and the larger person does not\n<image1>: the larger person wears yellow and the smaller person does not", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a big cat is next to a small dog\"\n2) \"a small cat is next to a big dog\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a small cat is next to a big dog\n<image1>: a big cat is next to a small dog", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the heavy oncoming traffic is contrasted with the light outgoing traffic\"\n2) \"the light oncoming traffic is contrasted with the heavy outgoing traffic\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the light oncoming traffic is contrasted with the heavy outgoing traffic\n<image1>: the heavy oncoming traffic is contrasted with the light outgoing traffic", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a hot drink on a cold day\"\n2) \"a cold drink on a hot day\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a cold drink on a hot day\n<image1>: a hot drink on a cold day", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a person is drinking later and eating now\"\n2) \"a person is drinking now and eating later\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a person is drinking now and eating later\n<image1>: a person is drinking later and eating now", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the sailboat sails close but the beach is far away\"\n2) \"the sailboat sails far away but the beach is close\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the sailboat sails close but the beach is far away\n<image1>: the sailboat sails far away but the beach is close", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the passenger is riding attentive to the road and the driver is driving inattentive to the road\"\n2) \"the passenger is riding inattentive to the road and the driver is driving attentive to the road\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the passenger is riding attentive to the road and the driver is driving inattentive to the road\n<image1>: the passenger is riding inattentive to the road and the driver is driving attentive to the road", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"it wears a hat but the person doesn't\"\n2) \"the person wears a hat but it doesn't\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person wears a hat but it doesn't\n<image1>: it wears a hat but the person doesn't", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with green legs is running faster and the red legged one runs quite slowly\"\n2) \"the person with green legs is running quite slowly and the red legged one runs faster\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with green legs is running faster and the red legged one runs quite slowly\n<image1>: the person with green legs is running quite slowly and the red legged one runs faster", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a person is close to the water and in the sand\"\n2) \"a person is in the water and close to the sand\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a person is in the water and close to the sand\n<image1>: a person is close to the water and in the sand", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a frog is above rocks which are under the water\"\n2) \"a frog is under rocks which are above the water\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a frog is above rocks which are under the water\n<image1>: a frog is under rocks which are above the water", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"driving on the road and off the unpaved terrain\"\n2) \"driving off the road and on the unpaved terrain\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: driving on the road and off the unpaved terrain\n<image1>: driving off the road and on the unpaved terrain", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone is with a helicopter in a car\"\n2) \"someone is in a helicopter with a car\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone is in a helicopter with a car\n<image1>: someone is with a helicopter in a car", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the green one can fit inside the blue one and the blue one can fit outside the green one\"\n2) \"the green one can fit outside the blue one and the blue one can fit inside the green one\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the green one can fit outside the blue one and the blue one can fit inside the green one\n<image1>: the green one can fit inside the blue one and the blue one can fit outside the green one", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"jumping towards the photographer and away from the others\"\n2) \"jumping away from the photographer and towards the others\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: jumping towards the photographer and away from the others\n<image1>: jumping away from the photographer and towards the others", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the dough is eaten before someone bakes the dough\"\n2) \"someone bakes the dough before the dough is eaten\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone bakes the dough before the dough is eaten\n<image1>: the dough is eaten before someone bakes the dough", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"it hatched before it was eaten\"\n2) \"it was eaten before it hatched\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: it was eaten before it hatched\n<image1>: it hatched before it was eaten", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are two bananas and three apples\"\n2) \"there are three bananas and two apples\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are two bananas and three apples\n<image1>: there are three bananas and two apples", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone has loaded their plate with two of the toast slices and one of the eggs\"\n2) \"someone has loaded their plate with one of the toast slices and two of the eggs\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone has loaded their plate with one of the toast slices and two of the eggs\n<image1>: someone has loaded their plate with two of the toast slices and one of the eggs", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are three people and two windows\"\n2) \"there are two people and three windows\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are three people and two windows\n<image1>: there are two people and three windows", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are 5 stars on the right flag and 50 stars on the left flag\"\n2) \"there are 50 stars on the right flag and 5 stars on the left flag\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are 50 stars on the right flag and 5 stars on the left flag\n<image1>: there are 5 stars on the right flag and 50 stars on the left flag", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"six people wear blue shirts and no people wear white shirts\"\n2) \"no people wear blue shirts and six people wear white shirts\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: no people wear blue shirts and six people wear white shirts\n<image1>: six people wear blue shirts and no people wear white shirts", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person's leg is on the dog's torso\"\n2) \"the dog's leg is on the person's torso\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person's leg is on the dog's torso\n<image1>: the dog's leg is on the person's torso", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person's eyes are bigger than the animal's\"\n2) \"the animal's eyes are bigger than the person's\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person's eyes are bigger than the animal's\n<image1>: the animal's eyes are bigger than the person's", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the shorter person's arm is around the taller person's shoulder\"\n2) \"the taller person's arm is around the shorter person's shoulder\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the shorter person's arm is around the taller person's shoulder\n<image1>: the taller person's arm is around the shorter person's shoulder", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person in a brown coat looks back and the person in a black coat looks forward\"\n2) \"the person in a black coat looks back and the person in a brown coat looks forward\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person in a black coat looks back and the person in a brown coat looks forward\n<image1>: the person in a brown coat looks back and the person in a black coat looks forward", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"it ran away while they pursued\"\n2) \"they ran away while it pursued\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: it ran away while they pursued\n<image1>: they ran away while it pursued", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"one is in a boat and almost everyone is swimming\"\n2) \"almost everyone is in a boat and one is swimming\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: one is in a boat and almost everyone is swimming\n<image1>: almost everyone is in a boat and one is swimming", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with the darker shoes is holding an emoji and the person with the lighter shoes isn't\"\n2) \"the person with the lighter shoes is holding an emoji and the person with the darker shoes isn't\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the lighter shoes is holding an emoji and the person with the darker shoes isn't\n<image1>: the person with the darker shoes is holding an emoji and the person with the lighter shoes isn't", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person wearing neutral colors poses and the person wearing brighter colors takes a picture\"\n2) \"the person wearing brighter colors poses and the person wearing neutral colors takes a picture\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person wearing brighter colors poses and the person wearing neutral colors takes a picture\n<image1>: the person wearing neutral colors poses and the person wearing brighter colors takes a picture", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"it's a truck fire\"\n2) \"it's a fire truck\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: it's a fire truck\n<image1>: it's a truck fire", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"watching the present\"\n2) \"presenting the watch\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: watching the present\n<image1>: presenting the watch", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"rocking out on a walk\"\n2) \"walking out on a rock\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: rocking out on a walk\n<image1>: walking out on a rock", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the kid with the magnifying glass looks at them\"\n2) \"the kid looks at them with the magnifying glass\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the kid with the magnifying glass looks at them\n<image1>: the kid looks at them with the magnifying glass", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person is sitting with the kids\"\n2) \"the person with the kids is sitting\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the kids is sitting\n<image1>: the person is sitting with the kids", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the people are touching each other with shoes\"\n2) \"the people with shoes are touching each other\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the people with shoes are touching each other\n<image1>: the people are touching each other with shoes", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone is hurt by a tree with an apple\"\n2) \"someone with an apple is hurt by a tree\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone is hurt by a tree with an apple\n<image1>: someone with an apple is hurt by a tree", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person is eating the food that is on the table\"\n2) \"the person that is on the table is eating the food\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person that is on the table is eating the food\n<image1>: the person is eating the food that is on the table", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the watering can is larger than the pot\"\n2) \"the pot is larger than the watering can\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the watering can is larger than the pot\n<image1>: the pot is larger than the watering can", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"they hold a suitcase handle with their left hand and they hold a passport with their right hand\"\n2) \"they hold a passport with their left hand and they hold a suitcase handle with their right hand\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: they hold a suitcase handle with their left hand and they hold a passport with their right hand\n<image1>: they hold a passport with their left hand and they hold a suitcase handle with their right hand", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with the patterned shirt is curly-haired and the other is straight-haired\"\n2) \"the person with the patterned shirt is straight-haired and the other is curly-haired\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the patterned shirt is straight-haired and the other is curly-haired\n<image1>: the person with the patterned shirt is curly-haired and the other is straight-haired", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a diagram showing movement from right to left\"\n2) \"a diagram showing movement from left to right\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a diagram showing movement from right to left\n<image1>: a diagram showing movement from left to right", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the feet are too big for these flip flops\"\n2) \"the flip flops are too big for these feet\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the flip flops are too big for these feet\n<image1>: the feet are too big for these flip flops", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person who is wearing a crown is kissing a frog\"\n2) \"the person is kissing a frog who is wearing a crown\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person who is wearing a crown is kissing a frog\n<image1>: the person is kissing a frog who is wearing a crown", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a dog sitting on a couch with a person lying on the floor\"\n2) \"a person lying on a couch with a dog sitting on the floor\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a dog sitting on a couch with a person lying on the floor\n<image1>: a person lying on a couch with a dog sitting on the floor", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"The caged bird has an unopened cage door\"\n2) \"The uncaged bird has an opened cage door\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: The caged bird has an unopened cage door\n<image1>: The uncaged bird has an opened cage door", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"meat with potatoes\"\n2) \"potatoes with meat\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: meat with potatoes\n<image1>: potatoes with meat", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person in pink was not close to winning\"\n2) \"the person in pink was close to not winning\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person in pink was not close to winning\n<image1>: the person in pink was close to not winning", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"two humans and one wheel\"\n2) \"two wheels and one human\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: two humans and one wheel\n<image1>: two wheels and one human", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"lines on clothing\"\n2) \"clothing on lines\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: clothing on lines\n<image1>: lines on clothing", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone talks on the phone happily while another person sits angrily\"\n2) \"someone talks on the phone angrily while another person sits happily\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone talks on the phone angrily while another person sits happily\n<image1>: someone talks on the phone happily while another person sits angrily", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"two boats and one direction down the stream\"\n2) \"one boat and two directions down the stream\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: one boat and two directions down the stream\n<image1>: two boats and one direction down the stream", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"The dog rides with a visible tongue out\"\n2) \"The dog rides without a visible tongue\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: The dog rides with a visible tongue out\n<image1>: The dog rides without a visible tongue", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"handing a hammer\"\n2) \"hammering a hand\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: hammering a hand\n<image1>: handing a hammer", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"shedding its leaves\"\n2) \"leaves its shedding\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: leaves its shedding\n<image1>: shedding its leaves", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"they are using bait but still got no fish\"\n2) \"they are using no bait but still got fish\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: they are using bait but still got no fish\n<image1>: they are using no bait but still got fish", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone on the ground is spraying water towards a vehicle\"\n2) \"someone is on a vehicle spraying water towards the ground\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone is on a vehicle spraying water towards the ground\n<image1>: someone on the ground is spraying water towards a vehicle", "role": "assistant" } ]
[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the dog bites what someone would normally wear as a hat\"\n2) \"the dog wears as a hat what someone would normally bite\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the dog wears as a hat what someone would normally bite\n<image1>: the dog bites what someone would normally wear as a hat", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a young person kisses an old person\"\n2) \"an old person kisses a young person\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a young person kisses an old person\n<image1>: an old person kisses a young person", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the shorter person hugs the taller person\"\n2) \"the taller person hugs the shorter person\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the shorter person hugs the taller person\n<image1>: the taller person hugs the shorter person", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the unmasked wrestler hits the masked wrestler\"\n2) \"the masked wrestler hits the unmasked wrestler\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the unmasked wrestler hits the masked wrestler\n<image1>: the masked wrestler hits the unmasked wrestler", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"an animal watches a person\"\n2) \"a person watches an animal\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: an animal watches a person\n<image1>: a person watches an animal", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with earrings pays the person without earrings\"\n2) \"the person without earrings pays the person with earrings\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with earrings pays the person without earrings\n<image1>: the person without earrings pays the person with earrings", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a snake eats a bird\"\n2) \"a bird eats a snake\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a bird eats a snake\n<image1>: a snake eats a bird", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"another organism was harmed by a plant, and that plant broke the organism into pieces\"\n2) \"a plant was harmed by another organism, and that organism broke the plant into pieces\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: another organism was harmed by a plant, and that plant broke the organism into pieces\n<image1>: a plant was harmed by another organism, and that organism broke the plant into pieces", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"water is in a bottle\"\n2) \"a bottle is in water\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: water is in a bottle\n<image1>: a bottle is in water", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a tree smashed into a car\"\n2) \"a car smashed into a tree\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a car smashed into a tree\n<image1>: a tree smashed into a car", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"people fall on the leaves\"\n2) \"the leaves fall on people\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the leaves fall on people\n<image1>: people fall on the leaves", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the water rests below the sail\"\n2) \"the sail rests below the water\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the sail rests below the water\n<image1>: the water rests below the sail", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is someone below a table\"\n2) \"there is a table below someone\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is a table below someone\n<image1>: there is someone below a table", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the image shows a computer on top of books\"\n2) \"the image shows books on top of a computer\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the image shows a computer on top of books\n<image1>: the image shows books on top of a computer", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a plant with some caterpillars\"\n2) \"a caterpillar with some plants\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a caterpillar with some plants\n<image1>: a plant with some caterpillars", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is some grass in a mug\"\n2) \"there is a mug in some grass\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is some grass in a mug\n<image1>: there is a mug in some grass", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are more snowboarders than skiers\"\n2) \"there are more skiers than snowboarders\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are more skiers than snowboarders\n<image1>: there are more snowboarders than skiers", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are fewer spoons than forks\"\n2) \"there are fewer forks than spoons\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are fewer spoons than forks\n<image1>: there are fewer forks than spoons", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are more ladybugs than flowers\"\n2) \"there are more flowers than ladybugs\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are more ladybugs than flowers\n<image1>: there are more flowers than ladybugs", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is less milk than orange juice\"\n2) \"there is less orange juice than milk\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is less milk than orange juice\n<image1>: there is less orange juice than milk", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is more dirt than empty space in the jar\"\n2) \"there is more empty space than dirt in the jar\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is more empty space than dirt in the jar\n<image1>: there is more dirt than empty space in the jar", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there is a bit less white chocolate than milk chocolate\"\n2) \"there is a bit less milk chocolate than white chocolate\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there is a bit less milk chocolate than white chocolate\n<image1>: there is a bit less white chocolate than milk chocolate", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a person stands and a dog sits\"\n2) \"a person sits and a dog stands\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a person sits and a dog stands\n<image1>: a person stands and a dog sits", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"The person with more facial hair lies awake while the other sleeps\"\n2) \"The person with more facial hair sleeps while the other lies awake\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: The person with more facial hair lies awake while the other sleeps\n<image1>: The person with more facial hair sleeps while the other lies awake", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person in the middle is standing and some others are floating\"\n2) \"the person in the middle is floating and some others are standing\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person in the middle is standing and some others are floating\n<image1>: the person in the middle is floating and some others are standing", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person closer to the camera weightlifts and the person farther from the camera runs\"\n2) \"the person closer to the camera runs and the person farther from the camera weightlifts\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person closer to the camera weightlifts and the person farther from the camera runs\n<image1>: the person closer to the camera runs and the person farther from the camera weightlifts", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with the white collared shirt holds the plant while the other waters it\"\n2) \"the person with the white collared shirt waters the plant while the other holds it\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the white collared shirt waters the plant while the other holds it\n<image1>: the person with the white collared shirt holds the plant while the other waters it", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the taller person chops food and the shorter person eats food\"\n2) \"the taller person eats food and the shorter person chops food\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the taller person eats food and the shorter person chops food\n<image1>: the taller person chops food and the shorter person eats food", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with the ponytail packs stuff and other buys it\"\n2) \"the person with the ponytail buys stuff and other packs it\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the ponytail buys stuff and other packs it\n<image1>: the person with the ponytail packs stuff and other buys it", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a person walking in a field of red flowers with a yellow dress\"\n2) \"a person walking in a field of yellow flowers with a red dress\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a person walking in a field of yellow flowers with a red dress\n<image1>: a person walking in a field of red flowers with a yellow dress", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a white dog is on a brown couch\"\n2) \"a brown dog is on a white couch\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a white dog is on a brown couch\n<image1>: a brown dog is on a white couch", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a red bird is next to a blue berry\"\n2) \"a blue bird is next to a red berry\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a red bird is next to a blue berry\n<image1>: a blue bird is next to a red berry", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"The red book is above the yellow book and below the blue book\"\n2) \"The yellow book is above the blue book and below the red book\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: The yellow book is above the blue book and below the red book\n<image1>: The red book is above the yellow book and below the blue book", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"astronauts in blue suits with a red planet in the background\"\n2) \"astronauts in red suits with a blue planet in the background\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: astronauts in blue suits with a red planet in the background\n<image1>: astronauts in red suits with a blue planet in the background", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the square table has a round base\"\n2) \"the round table has a square base\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the square table has a round base\n<image1>: the round table has a square base", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"rectangular bushes are behind pointy bushes\"\n2) \"pointy bushes are behind rectangular bushes\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: rectangular bushes are behind pointy bushes\n<image1>: pointy bushes are behind rectangular bushes", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a rectangular sign is above a circular sign\"\n2) \"a circular sign is above a rectangular sign\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a rectangular sign is above a circular sign\n<image1>: a circular sign is above a rectangular sign", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a metal chess piece rests on wood objects\"\n2) \"a wood chess piece rests on metal objects\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a metal chess piece rests on wood objects\n<image1>: a wood chess piece rests on metal objects", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"soft shoes are on a smooth floor\"\n2) \"smooth shoes are on a soft floor\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: smooth shoes are on a soft floor\n<image1>: soft shoes are on a smooth floor", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the happy person is on the right and the sad person is on the left\"\n2) \"the sad person is on the right and the happy person is on the left\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the sad person is on the right and the happy person is on the left\n<image1>: the happy person is on the right and the sad person is on the left", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the hurt person is on the left and the helpful person is on the right\"\n2) \"the helpful person is on the left and the hurt person is on the right\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the hurt person is on the left and the helpful person is on the right\n<image1>: the helpful person is on the left and the hurt person is on the right", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the less content person has the healthier food and the more content person has the less healthy food\"\n2) \"the more content person has the healthier food and the less content person has the less healthy food\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the more content person has the healthier food and the less content person has the less healthy food\n<image1>: the less content person has the healthier food and the more content person has the less healthy food", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the orange lollipop is surprised and the red lollipop is sad\"\n2) \"the orange lollipop is sad and the red lollipop is surprised\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the orange lollipop is sad and the red lollipop is surprised\n<image1>: the orange lollipop is surprised and the red lollipop is sad", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the smaller person wears yellow and the larger person does not\"\n2) \"the larger person wears yellow and the smaller person does not\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the smaller person wears yellow and the larger person does not\n<image1>: the larger person wears yellow and the smaller person does not", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a big cat is next to a small dog\"\n2) \"a small cat is next to a big dog\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a small cat is next to a big dog\n<image1>: a big cat is next to a small dog", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the heavy oncoming traffic is contrasted with the light outgoing traffic\"\n2) \"the light oncoming traffic is contrasted with the heavy outgoing traffic\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the light oncoming traffic is contrasted with the heavy outgoing traffic\n<image1>: the heavy oncoming traffic is contrasted with the light outgoing traffic", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a hot drink on a cold day\"\n2) \"a cold drink on a hot day\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a cold drink on a hot day\n<image1>: a hot drink on a cold day", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a person is drinking later and eating now\"\n2) \"a person is drinking now and eating later\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a person is drinking now and eating later\n<image1>: a person is drinking later and eating now", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the sailboat sails close but the beach is far away\"\n2) \"the sailboat sails far away but the beach is close\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the sailboat sails close but the beach is far away\n<image1>: the sailboat sails far away but the beach is close", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the passenger is riding attentive to the road and the driver is driving inattentive to the road\"\n2) \"the passenger is riding inattentive to the road and the driver is driving attentive to the road\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the passenger is riding attentive to the road and the driver is driving inattentive to the road\n<image1>: the passenger is riding inattentive to the road and the driver is driving attentive to the road", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"it wears a hat but the person doesn't\"\n2) \"the person wears a hat but it doesn't\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person wears a hat but it doesn't\n<image1>: it wears a hat but the person doesn't", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with green legs is running faster and the red legged one runs quite slowly\"\n2) \"the person with green legs is running quite slowly and the red legged one runs faster\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with green legs is running faster and the red legged one runs quite slowly\n<image1>: the person with green legs is running quite slowly and the red legged one runs faster", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a person is close to the water and in the sand\"\n2) \"a person is in the water and close to the sand\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a person is in the water and close to the sand\n<image1>: a person is close to the water and in the sand", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a frog is above rocks which are under the water\"\n2) \"a frog is under rocks which are above the water\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a frog is above rocks which are under the water\n<image1>: a frog is under rocks which are above the water", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"driving on the road and off the unpaved terrain\"\n2) \"driving off the road and on the unpaved terrain\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: driving on the road and off the unpaved terrain\n<image1>: driving off the road and on the unpaved terrain", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone is with a helicopter in a car\"\n2) \"someone is in a helicopter with a car\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone is in a helicopter with a car\n<image1>: someone is with a helicopter in a car", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the green one can fit inside the blue one and the blue one can fit outside the green one\"\n2) \"the green one can fit outside the blue one and the blue one can fit inside the green one\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the green one can fit outside the blue one and the blue one can fit inside the green one\n<image1>: the green one can fit inside the blue one and the blue one can fit outside the green one", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"jumping towards the photographer and away from the others\"\n2) \"jumping away from the photographer and towards the others\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: jumping towards the photographer and away from the others\n<image1>: jumping away from the photographer and towards the others", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the dough is eaten before someone bakes the dough\"\n2) \"someone bakes the dough before the dough is eaten\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone bakes the dough before the dough is eaten\n<image1>: the dough is eaten before someone bakes the dough", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"it hatched before it was eaten\"\n2) \"it was eaten before it hatched\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: it was eaten before it hatched\n<image1>: it hatched before it was eaten", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are two bananas and three apples\"\n2) \"there are three bananas and two apples\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are two bananas and three apples\n<image1>: there are three bananas and two apples", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone has loaded their plate with two of the toast slices and one of the eggs\"\n2) \"someone has loaded their plate with one of the toast slices and two of the eggs\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone has loaded their plate with one of the toast slices and two of the eggs\n<image1>: someone has loaded their plate with two of the toast slices and one of the eggs", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are three people and two windows\"\n2) \"there are two people and three windows\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are three people and two windows\n<image1>: there are two people and three windows", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"there are 5 stars on the right flag and 50 stars on the left flag\"\n2) \"there are 50 stars on the right flag and 5 stars on the left flag\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: there are 50 stars on the right flag and 5 stars on the left flag\n<image1>: there are 5 stars on the right flag and 50 stars on the left flag", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"six people wear blue shirts and no people wear white shirts\"\n2) \"no people wear blue shirts and six people wear white shirts\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: no people wear blue shirts and six people wear white shirts\n<image1>: six people wear blue shirts and no people wear white shirts", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person's leg is on the dog's torso\"\n2) \"the dog's leg is on the person's torso\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person's leg is on the dog's torso\n<image1>: the dog's leg is on the person's torso", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person's eyes are bigger than the animal's\"\n2) \"the animal's eyes are bigger than the person's\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person's eyes are bigger than the animal's\n<image1>: the animal's eyes are bigger than the person's", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the shorter person's arm is around the taller person's shoulder\"\n2) \"the taller person's arm is around the shorter person's shoulder\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the shorter person's arm is around the taller person's shoulder\n<image1>: the taller person's arm is around the shorter person's shoulder", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person in a brown coat looks back and the person in a black coat looks forward\"\n2) \"the person in a black coat looks back and the person in a brown coat looks forward\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person in a black coat looks back and the person in a brown coat looks forward\n<image1>: the person in a brown coat looks back and the person in a black coat looks forward", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"it ran away while they pursued\"\n2) \"they ran away while it pursued\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: it ran away while they pursued\n<image1>: they ran away while it pursued", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"one is in a boat and almost everyone is swimming\"\n2) \"almost everyone is in a boat and one is swimming\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: one is in a boat and almost everyone is swimming\n<image1>: almost everyone is in a boat and one is swimming", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with the darker shoes is holding an emoji and the person with the lighter shoes isn't\"\n2) \"the person with the lighter shoes is holding an emoji and the person with the darker shoes isn't\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the lighter shoes is holding an emoji and the person with the darker shoes isn't\n<image1>: the person with the darker shoes is holding an emoji and the person with the lighter shoes isn't", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person wearing neutral colors poses and the person wearing brighter colors takes a picture\"\n2) \"the person wearing brighter colors poses and the person wearing neutral colors takes a picture\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person wearing brighter colors poses and the person wearing neutral colors takes a picture\n<image1>: the person wearing neutral colors poses and the person wearing brighter colors takes a picture", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"it's a truck fire\"\n2) \"it's a fire truck\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: it's a fire truck\n<image1>: it's a truck fire", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"watching the present\"\n2) \"presenting the watch\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: watching the present\n<image1>: presenting the watch", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"rocking out on a walk\"\n2) \"walking out on a rock\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: rocking out on a walk\n<image1>: walking out on a rock", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the kid with the magnifying glass looks at them\"\n2) \"the kid looks at them with the magnifying glass\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the kid with the magnifying glass looks at them\n<image1>: the kid looks at them with the magnifying glass", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person is sitting with the kids\"\n2) \"the person with the kids is sitting\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the kids is sitting\n<image1>: the person is sitting with the kids", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the people are touching each other with shoes\"\n2) \"the people with shoes are touching each other\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the people with shoes are touching each other\n<image1>: the people are touching each other with shoes", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone is hurt by a tree with an apple\"\n2) \"someone with an apple is hurt by a tree\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone is hurt by a tree with an apple\n<image1>: someone with an apple is hurt by a tree", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person is eating the food that is on the table\"\n2) \"the person that is on the table is eating the food\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person that is on the table is eating the food\n<image1>: the person is eating the food that is on the table", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the watering can is larger than the pot\"\n2) \"the pot is larger than the watering can\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the watering can is larger than the pot\n<image1>: the pot is larger than the watering can", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"they hold a suitcase handle with their left hand and they hold a passport with their right hand\"\n2) \"they hold a passport with their left hand and they hold a suitcase handle with their right hand\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: they hold a suitcase handle with their left hand and they hold a passport with their right hand\n<image1>: they hold a passport with their left hand and they hold a suitcase handle with their right hand", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person with the patterned shirt is curly-haired and the other is straight-haired\"\n2) \"the person with the patterned shirt is straight-haired and the other is curly-haired\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person with the patterned shirt is straight-haired and the other is curly-haired\n<image1>: the person with the patterned shirt is curly-haired and the other is straight-haired", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a diagram showing movement from right to left\"\n2) \"a diagram showing movement from left to right\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a diagram showing movement from right to left\n<image1>: a diagram showing movement from left to right", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the feet are too big for these flip flops\"\n2) \"the flip flops are too big for these feet\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the flip flops are too big for these feet\n<image1>: the feet are too big for these flip flops", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person who is wearing a crown is kissing a frog\"\n2) \"the person is kissing a frog who is wearing a crown\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person who is wearing a crown is kissing a frog\n<image1>: the person is kissing a frog who is wearing a crown", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"a dog sitting on a couch with a person lying on the floor\"\n2) \"a person lying on a couch with a dog sitting on the floor\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: a dog sitting on a couch with a person lying on the floor\n<image1>: a person lying on a couch with a dog sitting on the floor", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"The caged bird has an unopened cage door\"\n2) \"The uncaged bird has an opened cage door\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: The caged bird has an unopened cage door\n<image1>: The uncaged bird has an opened cage door", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"meat with potatoes\"\n2) \"potatoes with meat\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: meat with potatoes\n<image1>: potatoes with meat", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the person in pink was not close to winning\"\n2) \"the person in pink was close to not winning\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the person in pink was not close to winning\n<image1>: the person in pink was close to not winning", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"two humans and one wheel\"\n2) \"two wheels and one human\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: two humans and one wheel\n<image1>: two wheels and one human", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"lines on clothing\"\n2) \"clothing on lines\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: clothing on lines\n<image1>: lines on clothing", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone talks on the phone happily while another person sits angrily\"\n2) \"someone talks on the phone angrily while another person sits happily\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone talks on the phone angrily while another person sits happily\n<image1>: someone talks on the phone happily while another person sits angrily", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"two boats and one direction down the stream\"\n2) \"one boat and two directions down the stream\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: one boat and two directions down the stream\n<image1>: two boats and one direction down the stream", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"The dog rides with a visible tongue out\"\n2) \"The dog rides without a visible tongue\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: The dog rides with a visible tongue out\n<image1>: The dog rides without a visible tongue", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"handing a hammer\"\n2) \"hammering a hand\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: hammering a hand\n<image1>: handing a hammer", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"shedding its leaves\"\n2) \"leaves its shedding\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: leaves its shedding\n<image1>: shedding its leaves", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"they are using bait but still got no fish\"\n2) \"they are using no bait but still got fish\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: they are using bait but still got no fish\n<image1>: they are using no bait but still got fish", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"someone on the ground is spraying water towards a vehicle\"\n2) \"someone is on a vehicle spraying water towards the ground\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: someone is on a vehicle spraying water towards the ground\n<image1>: someone on the ground is spraying water towards a vehicle", "role": "assistant" } ]

[ { "content": "There are two images: <image0> and <image1>.\nYou are given two candidate captions:\n1) \"the dog bites what someone would normally wear as a hat\"\n2) \"the dog wears as a hat what someone would normally bite\"\n\nAssign the correct caption to each image. Reply EXACTLY in this format:\n<image0>: <caption>\n<image1>: <caption>", "role": "user" }, { "content": "<image0>: the dog wears as a hat what someone would normally bite\n<image1>: the dog bites what someone would normally wear as a hat", "role": "assistant" } ]