Datasets:

---

zechen-nlp

/

star_bench

like 0

{ "choice": [ "Take the paper/notebook.", "Put down the cup/glass/bottle.", "Put down the box.", "Sit at the table." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after opening the refrigerator?\nOptions:\nA. Take the paper/notebook.\nB. Put down the cup/glass/bottle.\nC. Put down the box.\nD. Sit at the table.\nAnswer:", "A. Take the paper/notebook." ] }

{ "choice": [ "Put down the blanket.", "Throw the pillow.", "Put down the food.", "Open the book." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after turning off the light?\nOptions:\nA. Put down the blanket.\nB. Throw the pillow.\nC. Put down the food.\nD. Open the book.\nAnswer:", "A. Put down the blanket." ] }

{ "choice": [ "Wash the mirror.", "Open the book.", "Close the laptop.", "Throw the box." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after eating the sandwich?\nOptions:\nA. Wash the mirror.\nB. Open the book.\nC. Close the laptop.\nD. Throw the box.\nAnswer:", "B. Open the book." ] }

{ "choice": [ "Close the book.", "Hold the picture.", "Open the book.", "Take the food." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the sandwich?\nOptions:\nA. Close the book.\nB. Hold the picture.\nC. Open the book.\nD. Take the food.\nAnswer:", "C. Open the book." ] }

{ "choice": [ "Tidy up the table.", "Take the bag.", "Sit at the table.", "Put down the blanket." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after opening the box?\nOptions:\nA. Tidy up the table.\nB. Take the bag.\nC. Sit at the table.\nD. Put down the blanket.\nAnswer:", "C. Sit at the table." ] }

{ "choice": [ "Sit at the table.", "Wash the mirror.", "Close the closet/cabinet.", "Put down the box." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after opening the closet/cabinet?\nOptions:\nA. Sit at the table.\nB. Wash the mirror.\nC. Close the closet/cabinet.\nD. Put down the box.\nAnswer:", "D. Put down the box." ] }

{ "choice": [ "Sit at the table.", "Put down the food.", "Put down the pillow.", "Hold the mirror." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting the groceries?\nOptions:\nA. Sit at the table.\nB. Put down the food.\nC. Put down the pillow.\nD. Hold the mirror.\nAnswer:", "A. Sit at the table." ] }

{ "choice": [ "Take the paper/notebook.", "Close the laptop.", "Hold the food.", "Sit at the table." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the box?\nOptions:\nA. Take the paper/notebook.\nB. Close the laptop.\nC. Hold the food.\nD. Sit at the table.\nAnswer:", "D. Sit at the table." ] }

{ "choice": [ "Close the book.", "Sit at the table.", "Take the pillow.", "Put down the laptop." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after grasping onto the doorknob?\nOptions:\nA. Close the book.\nB. Sit at the table.\nC. Take the pillow.\nD. Put down the laptop.\nAnswer:", "B. Sit at the table." ] }

{ "choice": [ "Take the shoe.", "Hold the phone/camera.", "Take the food.", "Hold the dish." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after grasping onto the doorknob?\nOptions:\nA. Take the shoe.\nB. Hold the phone/camera.\nC. Take the food.\nD. Hold the dish.\nAnswer:", "C. Take the food." ] }

{ "choice": [ "Close the door.", "Take the dish.", "Tidy up the blanket.", "Sit at the table." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after grasping onto the doorknob?\nOptions:\nA. Close the door.\nB. Take the dish.\nC. Tidy up the blanket.\nD. Sit at the table.\nAnswer:", "A. Close the door." ] }

{ "choice": [ "Tidy up the clothes.", "Take the phone/camera.", "Put down the dish.", "Wash the window." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after drinking from the cup/glass/bottle?\nOptions:\nA. Tidy up the clothes.\nB. Take the phone/camera.\nC. Put down the dish.\nD. Wash the window.\nAnswer:", "D. Wash the window." ] }

{ "choice": [ "Sit at the table.", "Put down the towel.", "Put down the cup/glass/bottle.", "Sit on the bed." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the pillow?\nOptions:\nA. Sit at the table.\nB. Put down the towel.\nC. Put down the cup/glass/bottle.\nD. Sit on the bed.\nAnswer:", "A. Sit at the table." ] }

{ "choice": [ "Open the door.", "Put down the pillow.", "Throw the book.", "Close the closet/cabinet." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Open the door.\nB. Put down the pillow.\nC. Throw the book.\nD. Close the closet/cabinet.\nAnswer:", "B. Put down the pillow." ] }

{ "choice": [ "Throw the clothes.", "Sit at the table.", "Close the door.", "Put down the paper/notebook." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Throw the clothes.\nB. Sit at the table.\nC. Close the door.\nD. Put down the paper/notebook.\nAnswer:", "B. Sit at the table." ] }

{ "choice": [ "Eat the sandwich.", "Take the food.", "Close the book.", "Sit at the table." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after holding the blanket?\nOptions:\nA. Eat the sandwich.\nB. Take the food.\nC. Close the book.\nD. Sit at the table.\nAnswer:", "D. Sit at the table." ] }

{ "choice": [ "Sit on the bed.", "Put down the pillow.", "Hold the laptop.", "Close the door." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the blanket?\nOptions:\nA. Sit on the bed.\nB. Put down the pillow.\nC. Hold the laptop.\nD. Close the door.\nAnswer:", "B. Put down the pillow." ] }

{ "choice": [ "Throw the broom.", "Sit at the table.", "Put down the cup/glass/bottle.", "Put down the laptop." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the blanket?\nOptions:\nA. Throw the broom.\nB. Sit at the table.\nC. Put down the cup/glass/bottle.\nD. Put down the laptop.\nAnswer:", "B. Sit at the table." ] }

{ "choice": [ "Take the shoe.", "Close the door.", "Open the bag.", "Eat the medicine." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the cup/glass/bottle?\nOptions:\nA. Take the shoe.\nB. Close the door.\nC. Open the bag.\nD. Eat the medicine.\nAnswer:", "B. Close the door." ] }

{ "choice": [ "Put down the book.", "Open the door.", "Hold the clothes.", "Take the cup/glass/bottle." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the door?\nOptions:\nA. Put down the book.\nB. Open the door.\nC. Hold the clothes.\nD. Take the cup/glass/bottle.\nAnswer:", "D. Take the cup/glass/bottle." ] }

{ "choice": [ "Tidy up the blanket.", "Hold the phone/camera.", "Put down the towel.", "Put down the picture." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after opening the door?\nOptions:\nA. Tidy up the blanket.\nB. Hold the phone/camera.\nC. Put down the towel.\nD. Put down the picture.\nAnswer:", "A. Tidy up the blanket." ] }

{ "choice": [ "Put down the clothes.", "Hold the medicine.", "Hold the mirror.", "Tidy up the blanket." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the bag?\nOptions:\nA. Put down the clothes.\nB. Hold the medicine.\nC. Hold the mirror.\nD. Tidy up the blanket.\nAnswer:", "D. Tidy up the blanket." ] }

{ "choice": [ "Tidy up the blanket.", "Put down the clothes.", "Throw the clothes.", "Take the sandwich." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Tidy up the blanket.\nB. Put down the clothes.\nC. Throw the clothes.\nD. Take the sandwich.\nAnswer:", "A. Tidy up the blanket." ] }

{ "choice": [ "Tidy up the blanket.", "Throw the bag.", "Wash the dish.", "Hold the phone/camera." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after grasping onto the doorknob?\nOptions:\nA. Tidy up the blanket.\nB. Throw the bag.\nC. Wash the dish.\nD. Hold the phone/camera.\nAnswer:", "A. Tidy up the blanket." ] }

{ "choice": [ "Take the towel.", "Put down the food.", "Lie on the bed.", "Open the bag." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after holding the food?\nOptions:\nA. Take the towel.\nB. Put down the food.\nC. Lie on the bed.\nD. Open the bag.\nAnswer:", "D. Open the bag." ] }

{ "choice": [ "Throw the clothes.", "Throw the blanket.", "Put down the food.", "Lie on the sofa/couch." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the phone/camera?\nOptions:\nA. Throw the clothes.\nB. Throw the blanket.\nC. Put down the food.\nD. Lie on the sofa/couch.\nAnswer:", "B. Throw the blanket." ] }

{ "choice": [ "Tidy up the closet/cabinet.", "Throw the shoe.", "Put down the dish.", "Hold the pillow." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after drinking from the cup/glass/bottle?\nOptions:\nA. Tidy up the closet/cabinet.\nB. Throw the shoe.\nC. Put down the dish.\nD. Hold the pillow.\nAnswer:", "D. Hold the pillow." ] }

{ "choice": [ "Take the paper/notebook.", "Throw the clothes.", "Eat the sandwich.", "Hold the pillow." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the dish?\nOptions:\nA. Take the paper/notebook.\nB. Throw the clothes.\nC. Eat the sandwich.\nD. Hold the pillow.\nAnswer:", "D. Hold the pillow." ] }

{ "choice": [ "Throw the pillow.", "Hold the pillow.", "Put down the cup/glass/bottle.", "Close the box." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the dish?\nOptions:\nA. Throw the pillow.\nB. Hold the pillow.\nC. Put down the cup/glass/bottle.\nD. Close the box.\nAnswer:", "B. Hold the pillow." ] }

{ "choice": [ "Throw the shoe.", "Put down the clothes.", "Close the door.", "Tidy up the towel." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Throw the shoe.\nB. Put down the clothes.\nC. Close the door.\nD. Tidy up the towel.\nAnswer:", "B. Put down the clothes." ] }

{ "choice": [ "Wash the table.", "Put down the cup/glass/bottle.", "Take the shoe.", "Hold the pillow." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the dish?\nOptions:\nA. Wash the table.\nB. Put down the cup/glass/bottle.\nC. Take the shoe.\nD. Hold the pillow.\nAnswer:", "B. Put down the cup/glass/bottle." ] }

{ "choice": [ "Sit at the table.", "Put down the food.", "Throw the blanket.", "Open the door." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the closet/cabinet?\nOptions:\nA. Sit at the table.\nB. Put down the food.\nC. Throw the blanket.\nD. Open the door.\nAnswer:", "C. Throw the blanket." ] }

{ "choice": [ "Sit at the table.", "Take the bag.", "Tidy up the closet/cabinet.", "Throw the blanket." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the closet/cabinet?\nOptions:\nA. Sit at the table.\nB. Take the bag.\nC. Tidy up the closet/cabinet.\nD. Throw the blanket.\nAnswer:", "D. Throw the blanket." ] }

{ "choice": [ "Open the window.", "Put down the shoe.", "Take the towel.", "Take the paper/notebook." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the closet/cabinet?\nOptions:\nA. Open the window.\nB. Put down the shoe.\nC. Take the towel.\nD. Take the paper/notebook.\nAnswer:", "B. Put down the shoe." ] }

{ "choice": [ "Hold the laptop.", "Throw the pillow.", "Put down the bag.", "Sit on the sofa/couch." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after holding the pillow?\nOptions:\nA. Hold the laptop.\nB. Throw the pillow.\nC. Put down the bag.\nD. Sit on the sofa/couch.\nAnswer:", "D. Sit on the sofa/couch." ] }

{ "choice": [ "Sit at the table.", "Wash the table.", "Close the door.", "Hold the phone/camera." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after holding the clothes?\nOptions:\nA. Sit at the table.\nB. Wash the table.\nC. Close the door.\nD. Hold the phone/camera.\nAnswer:", "A. Sit at the table." ] }

{ "choice": [ "Sit at the table.", "Eat the sandwich.", "Throw the broom.", "Put down the bag." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the clothes?\nOptions:\nA. Sit at the table.\nB. Eat the sandwich.\nC. Throw the broom.\nD. Put down the bag.\nAnswer:", "A. Sit at the table." ] }

{ "choice": [ "Put down the cup/glass/bottle.", "Open the box.", "Sit at the table.", "Take the paper/notebook." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Put down the cup/glass/bottle.\nB. Open the box.\nC. Sit at the table.\nD. Take the paper/notebook.\nAnswer:", "C. Sit at the table." ] }

{ "choice": [ "Take the box.", "Close the book.", "Put down the clothes.", "Put down the laptop." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Take the box.\nB. Close the book.\nC. Put down the clothes.\nD. Put down the laptop.\nAnswer:", "C. Put down the clothes." ] }

{ "choice": [ "Sit at the table.", "Open the box.", "Tidy up the towel.", "Throw the book." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after grasping onto the doorknob?\nOptions:\nA. Sit at the table.\nB. Open the box.\nC. Tidy up the towel.\nD. Throw the book.\nAnswer:", "A. Sit at the table." ] }

{ "choice": [ "Wash the clothes.", "Throw the shoe.", "Put down the cup/glass/bottle.", "Sit at the table." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the door?\nOptions:\nA. Wash the clothes.\nB. Throw the shoe.\nC. Put down the cup/glass/bottle.\nD. Sit at the table.\nAnswer:", "D. Sit at the table." ] }

{ "choice": [ "Put down the clothes.", "Take the paper/notebook.", "Take the phone/camera.", "Put down the picture." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the door?\nOptions:\nA. Put down the clothes.\nB. Take the paper/notebook.\nC. Take the phone/camera.\nD. Put down the picture.\nAnswer:", "A. Put down the clothes." ] }

{ "choice": [ "Put down the clothes.", "Throw the blanket.", "Put down the sandwich.", "Hold the shoe." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after opening the door?\nOptions:\nA. Put down the clothes.\nB. Throw the blanket.\nC. Put down the sandwich.\nD. Hold the shoe.\nAnswer:", "A. Put down the clothes." ] }

{ "choice": [ "Close the refrigerator.", "Open the book.", "Open the door.", "Put down the blanket." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Close the refrigerator.\nB. Open the book.\nC. Open the door.\nD. Put down the blanket.\nAnswer:", "C. Open the door." ] }

{ "choice": [ "Close the book.", "Take the book.", "Take the phone/camera.", "Put down the shoe." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after lying on the floor?\nOptions:\nA. Close the book.\nB. Take the book.\nC. Take the phone/camera.\nD. Put down the shoe.\nAnswer:", "C. Take the phone/camera." ] }

{ "choice": [ "Wash the table.", "Open the closet/cabinet.", "Tidy up the table.", "Take the phone/camera." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting on the floor?\nOptions:\nA. Wash the table.\nB. Open the closet/cabinet.\nC. Tidy up the table.\nD. Take the phone/camera.\nAnswer:", "D. Take the phone/camera." ] }

{ "choice": [ "Take the book.", "Hold the picture.", "Hold the phone/camera.", "Hold the box." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting at the table?\nOptions:\nA. Take the book.\nB. Hold the picture.\nC. Hold the phone/camera.\nD. Hold the box.\nAnswer:", "C. Hold the phone/camera." ] }

{ "choice": [ "Put down the dish.", "Take the cup/glass/bottle.", "Put down the food.", "Hold the phone/camera." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the cup/glass/bottle?\nOptions:\nA. Put down the dish.\nB. Take the cup/glass/bottle.\nC. Put down the food.\nD. Hold the phone/camera.\nAnswer:", "D. Hold the phone/camera." ] }

{ "choice": [ "Open the closet/cabinet.", "Throw the food.", "Sit at the table.", "Throw the blanket." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after opening the closet/cabinet?\nOptions:\nA. Open the closet/cabinet.\nB. Throw the food.\nC. Sit at the table.\nD. Throw the blanket.\nAnswer:", "D. Throw the blanket." ] }

{ "choice": [ "Put down the towel.", "Wash the table.", "Open the bag.", "Take the phone/camera." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Put down the towel.\nB. Wash the table.\nC. Open the bag.\nD. Take the phone/camera.\nAnswer:", "A. Put down the towel." ] }

{ "choice": [ "Put down the cup/glass/bottle.", "Open the closet/cabinet.", "Throw the box.", "Sit at the table." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after holding the food?\nOptions:\nA. Put down the cup/glass/bottle.\nB. Open the closet/cabinet.\nC. Throw the box.\nD. Sit at the table.\nAnswer:", "D. Sit at the table." ] }

{ "choice": [ "Put down the towel.", "Open the closet/cabinet.", "Sit at the table.", "Take the phone/camera." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the closet/cabinet?\nOptions:\nA. Put down the towel.\nB. Open the closet/cabinet.\nC. Sit at the table.\nD. Take the phone/camera.\nAnswer:", "C. Sit at the table." ] }

{ "choice": [ "Sit at the table.", "Throw the clothes.", "Hold the pillow.", "Put down the dish." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the sandwich?\nOptions:\nA. Sit at the table.\nB. Throw the clothes.\nC. Hold the pillow.\nD. Put down the dish.\nAnswer:", "A. Sit at the table." ] }

{ "choice": [ "Put down the laptop.", "Put down the cup/glass/bottle.", "Open the bag.", "Tidy up the blanket." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Put down the laptop.\nB. Put down the cup/glass/bottle.\nC. Open the bag.\nD. Tidy up the blanket.\nAnswer:", "A. Put down the laptop." ] }

{ "choice": [ "Open the door.", "Put down the cup/glass/bottle.", "Put down the paper/notebook.", "Sit at the table." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after lying on the bed?\nOptions:\nA. Open the door.\nB. Put down the cup/glass/bottle.\nC. Put down the paper/notebook.\nD. Sit at the table.\nAnswer:", "D. Sit at the table." ] }

{ "choice": [ "Take the book.", "Throw the clothes.", "Open the closet/cabinet.", "Put down the food." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the sandwich?\nOptions:\nA. Take the book.\nB. Throw the clothes.\nC. Open the closet/cabinet.\nD. Put down the food.\nAnswer:", "D. Put down the food." ] }

{ "choice": [ "Wash the dish.", "Hold the food.", "Put down the dish.", "Take the sandwich." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting on the chair?\nOptions:\nA. Wash the dish.\nB. Hold the food.\nC. Put down the dish.\nD. Take the sandwich.\nAnswer:", "B. Hold the food." ] }

{ "choice": [ "Put down the dish.", "Put down the paper/notebook.", "Hold the pillow.", "Take the pillow." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the food?\nOptions:\nA. Put down the dish.\nB. Put down the paper/notebook.\nC. Hold the pillow.\nD. Take the pillow.\nAnswer:", "A. Put down the dish." ] }

{ "choice": [ "Wash the clothes.", "Sit at the table.", "Lie on the bed.", "Close the door." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Wash the clothes.\nB. Sit at the table.\nC. Lie on the bed.\nD. Close the door.\nAnswer:", "D. Close the door." ] }

{ "choice": [ "Close the door.", "Eat the sandwich.", "Put down the clothes.", "Open the door." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after grasping onto the doorknob?\nOptions:\nA. Close the door.\nB. Eat the sandwich.\nC. Put down the clothes.\nD. Open the door.\nAnswer:", "A. Close the door." ] }

{ "choice": [ "Put down the paper/notebook.", "Sit at the table.", "Close the book.", "Hold the pillow." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the book?\nOptions:\nA. Put down the paper/notebook.\nB. Sit at the table.\nC. Close the book.\nD. Hold the pillow.\nAnswer:", "A. Put down the paper/notebook." ] }

{ "choice": [ "Eat the sandwich.", "Put down the clothes.", "Tidy up the table.", "Hold the mirror." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting at the table?\nOptions:\nA. Eat the sandwich.\nB. Put down the clothes.\nC. Tidy up the table.\nD. Hold the mirror.\nAnswer:", "A. Eat the sandwich." ] }

{ "choice": [ "Throw the bag.", "Take the box.", "Put down the bag.", "Tidy up the table." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting at the table?\nOptions:\nA. Throw the bag.\nB. Take the box.\nC. Put down the bag.\nD. Tidy up the table.\nAnswer:", "A. Throw the bag." ] }

{ "choice": [ "Throw the food.", "Open the box.", "Hold the pillow.", "Put down the pillow." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after holding the laptop?\nOptions:\nA. Throw the food.\nB. Open the box.\nC. Hold the pillow.\nD. Put down the pillow.\nAnswer:", "D. Put down the pillow." ] }

{ "choice": [ "Hold the pillow.", "Close the laptop.", "Put down the pillow.", "Put down the picture." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting on the sofa/couch?\nOptions:\nA. Hold the pillow.\nB. Close the laptop.\nC. Put down the pillow.\nD. Put down the picture.\nAnswer:", "C. Put down the pillow." ] }

{ "choice": [ "Put down the picture.", "Put down the clothes.", "Take the paper/notebook.", "Hold the mirror." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Put down the picture.\nB. Put down the clothes.\nC. Take the paper/notebook.\nD. Hold the mirror.\nAnswer:", "A. Put down the picture." ] }

{ "choice": [ "Put down the towel.", "Put down the box.", "Open the bag.", "Put down the shoe." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after opening the closet/cabinet?\nOptions:\nA. Put down the towel.\nB. Put down the box.\nC. Open the bag.\nD. Put down the shoe.\nAnswer:", "A. Put down the towel." ] }

{ "choice": [ "Hold the shoe.", "Open the door.", "Put down the paper/notebook.", "Put down the book." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Hold the shoe.\nB. Open the door.\nC. Put down the paper/notebook.\nD. Put down the book.\nAnswer:", "A. Hold the shoe." ] }

{ "choice": [ "Put down the dish.", "Put down the towel.", "Open the bag.", "Sit at the table." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after grasping onto the doorknob?\nOptions:\nA. Put down the dish.\nB. Put down the towel.\nC. Open the bag.\nD. Sit at the table.\nAnswer:", "B. Put down the towel." ] }

{ "choice": [ "Sit on the bed.", "Put down the laptop.", "Put down the dish.", "Close the box." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting on the chair?\nOptions:\nA. Sit on the bed.\nB. Put down the laptop.\nC. Put down the dish.\nD. Close the box.\nAnswer:", "D. Close the box." ] }

{ "choice": [ "Hold the book.", "Put down the dish.", "Hold the laptop.", "Put down the phone/camera." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after lying on the floor?\nOptions:\nA. Hold the book.\nB. Put down the dish.\nC. Hold the laptop.\nD. Put down the phone/camera.\nAnswer:", "A. Hold the book." ] }

{ "choice": [ "Open the laptop.", "Close the door.", "Wash the mirror.", "Put down the dish." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after grasping onto the doorknob?\nOptions:\nA. Open the laptop.\nB. Close the door.\nC. Wash the mirror.\nD. Put down the dish.\nAnswer:", "B. Close the door." ] }

{ "choice": [ "Close the laptop.", "Put down the food.", "Take the phone/camera.", "Put down the cup/glass/bottle." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after watching outside the window?\nOptions:\nA. Close the laptop.\nB. Put down the food.\nC. Take the phone/camera.\nD. Put down the cup/glass/bottle.\nAnswer:", "D. Put down the cup/glass/bottle." ] }

{ "choice": [ "Close the door.", "Hold the box.", "Put down the cup/glass/bottle.", "Open the bag." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the door?\nOptions:\nA. Close the door.\nB. Hold the box.\nC. Put down the cup/glass/bottle.\nD. Open the bag.\nAnswer:", "C. Put down the cup/glass/bottle." ] }

{ "choice": [ "Sit on the sofa/couch.", "Put down the laptop.", "Sit at the table.", "Throw the shoe." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting on the sofa/couch?\nOptions:\nA. Sit on the sofa/couch.\nB. Put down the laptop.\nC. Sit at the table.\nD. Throw the shoe.\nAnswer:", "D. Throw the shoe." ] }

{ "choice": [ "Take the paper/notebook.", "Sit at the table.", "Hold the medicine.", "Throw the shoe." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the cup/glass/bottle?\nOptions:\nA. Take the paper/notebook.\nB. Sit at the table.\nC. Hold the medicine.\nD. Throw the shoe.\nAnswer:", "B. Sit at the table." ] }

{ "choice": [ "Put down the laptop.", "Take the sandwich.", "Put down the dish.", "Take the dish." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after holding the food?\nOptions:\nA. Put down the laptop.\nB. Take the sandwich.\nC. Put down the dish.\nD. Take the dish.\nAnswer:", "B. Take the sandwich." ] }

{ "choice": [ "Put down the cup/glass/bottle.", "Put down the clothes.", "Take the box.", "Throw the food." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting on the sofa/couch?\nOptions:\nA. Put down the cup/glass/bottle.\nB. Put down the clothes.\nC. Take the box.\nD. Throw the food.\nAnswer:", "B. Put down the clothes." ] }

{ "choice": [ "Take the cup/glass/bottle.", "Put down the clothes.", "Close the book.", "Put down the phone/camera." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting on the floor?\nOptions:\nA. Take the cup/glass/bottle.\nB. Put down the clothes.\nC. Close the book.\nD. Put down the phone/camera.\nAnswer:", "C. Close the book." ] }

{ "choice": [ "Sit at the table.", "Sit on the floor.", "Wash the table.", "Lie on the bed." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the shoe?\nOptions:\nA. Sit at the table.\nB. Sit on the floor.\nC. Wash the table.\nD. Lie on the bed.\nAnswer:", "A. Sit at the table." ] }

{ "choice": [ "Tidy up the closet/cabinet.", "Take the clothes.", "Sit at the table.", "Take the sandwich." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the shoe?\nOptions:\nA. Tidy up the closet/cabinet.\nB. Take the clothes.\nC. Sit at the table.\nD. Take the sandwich.\nAnswer:", "C. Sit at the table." ] }

{ "choice": [ "Hold the mirror.", "Put down the food.", "Hold the food.", "Open the closet/cabinet." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting the groceries?\nOptions:\nA. Hold the mirror.\nB. Put down the food.\nC. Hold the food.\nD. Open the closet/cabinet.\nAnswer:", "C. Hold the food." ] }

{ "choice": [ "Throw the towel.", "Sit at the table.", "Put down the clothes.", "Hold the food." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting the groceries?\nOptions:\nA. Throw the towel.\nB. Sit at the table.\nC. Put down the clothes.\nD. Hold the food.\nAnswer:", "B. Sit at the table." ] }

{ "choice": [ "Hold the mirror.", "Hold the food.", "Eat the medicine.", "Close the closet/cabinet." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the bag?\nOptions:\nA. Hold the mirror.\nB. Hold the food.\nC. Eat the medicine.\nD. Close the closet/cabinet.\nAnswer:", "B. Hold the food." ] }

{ "choice": [ "Close the refrigerator.", "Lie on the bed.", "Open the window.", "Put down the laptop." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the food?\nOptions:\nA. Close the refrigerator.\nB. Lie on the bed.\nC. Open the window.\nD. Put down the laptop.\nAnswer:", "A. Close the refrigerator." ] }

{ "choice": [ "Put down the food.", "Open the window.", "Put down the cup/glass/bottle.", "Wash the clothes." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after watching outside the window?\nOptions:\nA. Put down the food.\nB. Open the window.\nC. Put down the cup/glass/bottle.\nD. Wash the clothes.\nAnswer:", "A. Put down the food." ] }

{ "choice": [ "Take the clothes.", "Take the shoe.", "Throw the shoe.", "Hold the box." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting on the bed?\nOptions:\nA. Take the clothes.\nB. Take the shoe.\nC. Throw the shoe.\nD. Hold the box.\nAnswer:", "C. Throw the shoe." ] }

{ "choice": [ "Put down the blanket.", "Hold the box.", "Tidy up the table.", "Take the cup/glass/bottle." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the clothes?\nOptions:\nA. Put down the blanket.\nB. Hold the box.\nC. Tidy up the table.\nD. Take the cup/glass/bottle.\nAnswer:", "D. Take the cup/glass/bottle." ] }

{ "choice": [ "Take the cup/glass/bottle.", "Take the towel.", "Put down the blanket.", "Tidy up the clothes." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after throwing the clothes?\nOptions:\nA. Take the cup/glass/bottle.\nB. Take the towel.\nC. Put down the blanket.\nD. Tidy up the clothes.\nAnswer:", "A. Take the cup/glass/bottle." ] }

{ "choice": [ "Throw the blanket.", "Tidy up the table.", "Lie on the sofa/couch.", "Put down the laptop." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after lying on the sofa/couch?\nOptions:\nA. Throw the blanket.\nB. Tidy up the table.\nC. Lie on the sofa/couch.\nD. Put down the laptop.\nAnswer:", "B. Tidy up the table." ] }

{ "choice": [ "Take the laptop.", "Sit at the table.", "Hold the pillow.", "Put down the clothes." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after lying on the sofa/couch?\nOptions:\nA. Take the laptop.\nB. Sit at the table.\nC. Hold the pillow.\nD. Put down the clothes.\nAnswer:", "D. Put down the clothes." ] }

{ "choice": [ "Sit on the bed.", "Sit on the sofa/couch.", "Open the box.", "Put down the cup/glass/bottle." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after sitting at the table?\nOptions:\nA. Sit on the bed.\nB. Sit on the sofa/couch.\nC. Open the box.\nD. Put down the cup/glass/bottle.\nAnswer:", "B. Sit on the sofa/couch." ] }

{ "choice": [ "Hold the blanket.", "Throw the book.", "Close the door.", "Take the cup/glass/bottle." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after grasping onto the doorknob?\nOptions:\nA. Hold the blanket.\nB. Throw the book.\nC. Close the door.\nD. Take the cup/glass/bottle.\nAnswer:", "C. Close the door." ] }

{ "choice": [ "Close the door.", "Throw the book.", "Sit at the table.", "Hold the mirror." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Close the door.\nB. Throw the book.\nC. Sit at the table.\nD. Hold the mirror.\nAnswer:", "A. Close the door." ] }

{ "choice": [ "Take the blanket.", "Put down the dish.", "Throw the clothes.", "Throw the food." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after taking the sandwich?\nOptions:\nA. Take the blanket.\nB. Put down the dish.\nC. Throw the clothes.\nD. Throw the food.\nAnswer:", "B. Put down the dish." ] }

{ "choice": [ "Put down the bag.", "Throw the shoe.", "Take the shoe.", "Open the bag." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after closing the door?\nOptions:\nA. Put down the bag.\nB. Throw the shoe.\nC. Take the shoe.\nD. Open the bag.\nAnswer:", "C. Take the shoe." ] }

{ "choice": [ "Take the sandwich.", "Take the shoe.", "Tidy up the table.", "Put down the paper/notebook." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after walking through the doorway?\nOptions:\nA. Take the sandwich.\nB. Take the shoe.\nC. Tidy up the table.\nD. Put down the paper/notebook.\nAnswer:", "B. Take the shoe." ] }

{ "choice": [ "Hold the phone/camera.", "Take the dish.", "Take the pillow.", "Throw the food." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after holding the food?\nOptions:\nA. Hold the phone/camera.\nB. Take the dish.\nC. Take the pillow.\nD. Throw the food.\nAnswer:", "B. Take the dish." ] }

{ "choice": [ "Eat the sandwich.", "Hold the food.", "Hold the clothes.", "Take the cup/glass/bottle." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after putting down the food?\nOptions:\nA. Eat the sandwich.\nB. Hold the food.\nC. Hold the clothes.\nD. Take the cup/glass/bottle.\nAnswer:", "C. Hold the clothes." ] }

{ "choice": [ "Tidy up the table.", "Take the clothes.", "Put down the blanket.", "Close the door." ], "choice_id": [ 0, 1, 2, 3 ] }

{ "from": [ "human", "gpt" ], "value": [ "<image>\nAnalyze the given video clip and select the best answer for the following question.\nQuestion: What is the person able to do after watching the television?\nOptions:\nA. Tidy up the table.\nB. Take the clothes.\nC. Put down the blanket.\nD. Close the door.\nAnswer:", "B. Take the clothes." ] }