id
stringlengths
2
214
conversations
listlengths
2
42
data_source
stringclasses
54 values
cauldron/dvqa/images/dvqa_00099886.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Auto." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "90." }, { "from": "human", "value": "How many algorithms have accuracies higher than 80?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Is the accuracy of the algorithm baby larger than hurry?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm hurry?" }, { "from": "gpt", "value": "80." }, { "from": "human", "value": "What is the label of the first bar from the bottom?" }, { "from": "gpt", "value": "Stroke." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Four." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099891.png
[ { "from": "human", "value": "Which item sold the most units?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Hay." }, { "from": "human", "value": "Which item sold the least units?" }, { "from": "gpt", "value": "Sera." }, { "from": "human", "value": "How many units of the the most sold item were sold?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many units of the the least sold item were sold?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "How many more of the most sold item were sold compared to the least sold item?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many items sold more than 5 units?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "How many units of items hay and status were sold?" }, { "from": "gpt", "value": "12." }, { "from": "human", "value": "Did the item sera sold less units than hay?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many units of the item hay were sold?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Sera." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099895.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Sin." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Hall." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "How many objects are liked by more than 7 people?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "How many people prefer the objects heard or hall?" }, { "from": "gpt", "value": "11." }, { "from": "human", "value": "Is the object sin preferred by less people than heard?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many people prefer the object sin?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the label of the first bar from the bottom?" }, { "from": "gpt", "value": "Hall." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099896.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Train." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Winter." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many algorithms have accuracies lower than 1?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms bottom and gift?" }, { "from": "gpt", "value": "11." }, { "from": "human", "value": "Is the accuracy of the algorithm bottom smaller than winter?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the accuracy of the algorithm train?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the label of the second bar from the bottom?" }, { "from": "gpt", "value": "Train." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099911.png
[ { "from": "human", "value": "How many stacks of bars contain at least one element with value greater than 70?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which stack of bars contains the largest valued individual element in the whole chart?" }, { "from": "gpt", "value": "State." }, { "from": "human", "value": "Which stack of bars contains the smallest valued individual element in the whole chart?" }, { "from": "gpt", "value": "State." }, { "from": "human", "value": "What is the value of the largest individual element in the whole chart?" }, { "from": "gpt", "value": "90." }, { "from": "human", "value": "What is the value of the smallest individual element in the whole chart?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the sienna color represent?" }, { "from": "gpt", "value": "Work." }, { "from": "human", "value": "What is the value of peace in moment?" }, { "from": "gpt", "value": "30." }, { "from": "human", "value": "What is the label of the seventh stack of bars from the left?" }, { "from": "gpt", "value": "Forest." }, { "from": "human", "value": "What is the label of the second element from the bottom in each stack of bars?" }, { "from": "gpt", "value": "Work." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099913.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Humor." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Tour." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "60." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "50." }, { "from": "human", "value": "How many algorithms have accuracies lower than 60?" }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Is the accuracy of the algorithm tour larger than humor?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm shadow?" }, { "from": "gpt", "value": "50." }, { "from": "human", "value": "What is the label of the fifth bar from the left?" }, { "from": "gpt", "value": "Tour." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099914.png
[ { "from": "human", "value": "Which bar has the largest value?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Tax." }, { "from": "human", "value": "Which bar has the smallest value?" }, { "from": "gpt", "value": "Soul." }, { "from": "human", "value": "What is the value of the largest bar?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the value of the smallest bar?" }, { "from": "gpt", "value": "-4." }, { "from": "human", "value": "How many bars have values smaller than 2?" }, { "from": "gpt", "value": "Three." }, { "from": "human", "value": "Is the value of band larger than tax?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the value of soul?" }, { "from": "gpt", "value": "-4." }, { "from": "human", "value": "What is the label of the third bar from the bottom?" }, { "from": "gpt", "value": "Band." }, { "from": "human", "value": "Does the chart contain any negative values?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099917.png
[ { "from": "human", "value": "How many objects are preferred by more than 9 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Tire." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Tire." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Coat." }, { "from": "human", "value": "How many total people preferred the object tire across all the categories?" }, { "from": "gpt", "value": "11." }, { "from": "human", "value": "Is the object detail in the category merit preferred by more people than the object blue in the category court?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the lightseagreen color represent?" }, { "from": "gpt", "value": "Court." }, { "from": "human", "value": "How many people prefer the object floor in the category youth?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the third group of bars from the bottom?" }, { "from": "gpt", "value": "Tire." }, { "from": "human", "value": "What is the label of the third bar from the bottom in each group?" }, { "from": "gpt", "value": "Youth." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099927.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Family." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Sex." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "How many objects are liked by more than 5 people?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "How many people prefer the objects path or family?" }, { "from": "gpt", "value": "11." }, { "from": "human", "value": "Is the object family preferred by less people than mode?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many people prefer the object means?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "What is the label of the sixth bar from the left?" }, { "from": "gpt", "value": "Mode." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Seven." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099929.png
[ { "from": "human", "value": "Which item sold the most units?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Door." }, { "from": "human", "value": "Which item sold the least units?" }, { "from": "gpt", "value": "Back." }, { "from": "human", "value": "How many units of the the most sold item were sold?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many units of the the least sold item were sold?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "How many more of the most sold item were sold compared to the least sold item?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "How many items sold less than 4 units?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "How many units of items chief and role were sold?" }, { "from": "gpt", "value": "12." }, { "from": "human", "value": "Did the item door sold less units than bone?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many units of the item chief were sold?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the fifth bar from the left?" }, { "from": "gpt", "value": "Back." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099931.png
[ { "from": "human", "value": "Which bar has the largest value?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Knee." }, { "from": "human", "value": "Which bar has the smallest value?" }, { "from": "gpt", "value": "Flavor." }, { "from": "human", "value": "What is the value of the largest bar?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "What is the value of the smallest bar?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the difference between the largest and the smallest value in the chart?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "How many bars have values larger than 5?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "What is the sum of the values of flavor and effect?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "Is the value of bill smaller than knee?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the value of knee?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "What is the label of the first bar from the bottom?" }, { "from": "gpt", "value": "Knee." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099933.png
[ { "from": "human", "value": "How many objects are preferred by more than 6 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Box." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Deal." }, { "from": "human", "value": "How many people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Deal." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Box." }, { "from": "human", "value": "How many total people preferred the object deal across all the categories?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "Is the object box in the category handle preferred by less people than the object deal in the category parlor?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What category does the blue color represent?" }, { "from": "gpt", "value": "Handle." }, { "from": "human", "value": "How many people prefer the object box in the category handle?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the first group of bars from the left?" }, { "from": "gpt", "value": "Deal." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Parlor." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099940.png
[ { "from": "human", "value": "How many algorithms have accuracy higher than 2 in at least one dataset?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Seven." }, { "from": "human", "value": "Which algorithm has the smallest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Bulk." }, { "from": "human", "value": "Which algorithm has the largest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Factor." }, { "from": "human", "value": "What is the sum of accuracies of the algorithm tape for all the datasets?" }, { "from": "gpt", "value": "17." }, { "from": "human", "value": "Is the accuracy of the algorithm factor in the dataset page smaller than the accuracy of the algorithm critic in the dataset party?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What dataset does the steelblue color represent?" }, { "from": "gpt", "value": "Dark." }, { "from": "human", "value": "What is the accuracy of the algorithm output in the dataset dark?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the label of the third group of bars from the bottom?" }, { "from": "gpt", "value": "Remark." }, { "from": "human", "value": "What is the label of the second bar from the bottom in each group?" }, { "from": "gpt", "value": "Dark." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099943.png
[ { "from": "human", "value": "How many items sold more than 1 units in at least one store?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Six." }, { "from": "human", "value": "Which item sold the least number of units summed across all the stores?" }, { "from": "gpt", "value": "Cap." }, { "from": "human", "value": "Which item sold the most number of units summed across all the stores?" }, { "from": "gpt", "value": "Voting." }, { "from": "human", "value": "How many units of the item driver were sold across all the stores?" }, { "from": "gpt", "value": "26." }, { "from": "human", "value": "Did the item throat in the store duty sold smaller units than the item cell in the store cut?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What store does the darkturquoise color represent?" }, { "from": "gpt", "value": "Cut." }, { "from": "human", "value": "How many units of the item voting were sold in the store parlor?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the label of the third stack of bars from the bottom?" }, { "from": "gpt", "value": "Talk." }, { "from": "human", "value": "What is the label of the second element from the left in each stack of bars?" }, { "from": "gpt", "value": "Desire." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099950.png
[ { "from": "human", "value": "Which item sold the most units?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Work." }, { "from": "human", "value": "Which item sold the least units?" }, { "from": "gpt", "value": "Lobby." }, { "from": "human", "value": "How many units of the the most sold item were sold?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "How many units of the the least sold item were sold?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "How many more of the most sold item were sold compared to the least sold item?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "How many items sold more than 6 units?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "How many units of items work and knife were sold?" }, { "from": "gpt", "value": "15." }, { "from": "human", "value": "Did the item work sold more units than knife?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many units of the item knife were sold?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the first bar from the bottom?" }, { "from": "gpt", "value": "Work." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099953.png
[ { "from": "human", "value": "Which bar has the largest value?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Brand." }, { "from": "human", "value": "Which bar has the smallest value?" }, { "from": "gpt", "value": "Grade." }, { "from": "human", "value": "What is the value of the largest bar?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the value of the smallest bar?" }, { "from": "gpt", "value": "-4." }, { "from": "human", "value": "How many bars have values smaller than -4?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Is the value of grade smaller than brand?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the value of grade?" }, { "from": "gpt", "value": "-4." }, { "from": "human", "value": "What is the label of the first bar from the bottom?" }, { "from": "gpt", "value": "Hall." }, { "from": "human", "value": "Does the chart contain any negative values?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099961.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Reach." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Mood." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "How many objects are liked by less than 4 people?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "How many people prefer the objects year or mood?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "Is the object reach preferred by more people than mood?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many people prefer the object year?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "What is the label of the third bar from the left?" }, { "from": "gpt", "value": "Mood." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099963.png
[ { "from": "human", "value": "How many objects are preferred by less than 9 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Five." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Bench." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Party." }, { "from": "human", "value": "How many total people preferred the object sorry across all the categories?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Is the object sorry in the category return preferred by more people than the object party in the category finger?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What category does the steelblue color represent?" }, { "from": "gpt", "value": "Return." }, { "from": "human", "value": "How many people prefer the object vice in the category stroke?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the label of the second stack of bars from the left?" }, { "from": "gpt", "value": "Party." }, { "from": "human", "value": "What is the label of the first element from the bottom in each stack of bars?" }, { "from": "gpt", "value": "Return." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099966.png
[ { "from": "human", "value": "How many objects are preferred by less than 6 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Food." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Food." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Detail." }, { "from": "human", "value": "How many total people preferred the object tale across all the categories?" }, { "from": "gpt", "value": "11." }, { "from": "human", "value": "What category does the red color represent?" }, { "from": "gpt", "value": "Bird." }, { "from": "human", "value": "How many people prefer the object food in the category smell?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the first group of bars from the left?" }, { "from": "gpt", "value": "Tale." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Smell." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099972.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than 3?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Five." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "View." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Weight." }, { "from": "human", "value": "Is the value of line in fact smaller than the value of view in fish?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the darkturquoise color represent?" }, { "from": "gpt", "value": "Stream." }, { "from": "human", "value": "What is the value of fish in twenty?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the label of the sixth group of bars from the left?" }, { "from": "gpt", "value": "Twenty." }, { "from": "human", "value": "What is the label of the third bar from the left in each group?" }, { "from": "gpt", "value": "Trap." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many bars are there per group?" }, { "from": "gpt", "value": "Four." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099974.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Guard." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Bulk." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many objects are liked by more than 5 people?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "How many people prefer the objects guard or bulk?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Is the object guard preferred by more people than bulk?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many people prefer the object bulk?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the first bar from the bottom?" }, { "from": "gpt", "value": "Guard." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Three." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099976.png
[ { "from": "human", "value": "How many algorithms have accuracy lower than 8 in at least one dataset?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which algorithm has highest accuracy for any dataset?" }, { "from": "gpt", "value": "Birth." }, { "from": "human", "value": "Which algorithm has lowest accuracy for any dataset?" }, { "from": "gpt", "value": "Policy." }, { "from": "human", "value": "What is the highest accuracy reported in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the lowest accuracy reported in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which algorithm has the smallest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Policy." }, { "from": "human", "value": "Which algorithm has the largest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Birth." }, { "from": "human", "value": "What is the sum of accuracies of the algorithm policy for all the datasets?" }, { "from": "gpt", "value": "12." }, { "from": "human", "value": "Is the accuracy of the algorithm policy in the dataset choice larger than the accuracy of the algorithm birth in the dataset pack?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What dataset does the sienna color represent?" }, { "from": "gpt", "value": "Choice." }, { "from": "human", "value": "What is the accuracy of the algorithm birth in the dataset pack?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the label of the first group of bars from the left?" }, { "from": "gpt", "value": "Policy." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Choice." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099981.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than 40?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Three." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Female." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Is the value of golf in skill larger than the value of drunk in step?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the forestgreen color represent?" }, { "from": "gpt", "value": "Cellar." }, { "from": "human", "value": "What is the value of cellar in spirit?" }, { "from": "gpt", "value": "30." }, { "from": "human", "value": "What is the label of the sixth group of bars from the bottom?" }, { "from": "gpt", "value": "Spirit." }, { "from": "human", "value": "What is the label of the second bar from the bottom in each group?" }, { "from": "gpt", "value": "Skill." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many groups of bars are there?" }, { "from": "gpt", "value": "Six." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099992.png
[ { "from": "human", "value": "How many algorithms have accuracy higher than 2 in at least one dataset?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which algorithm has highest accuracy for any dataset?" }, { "from": "gpt", "value": "Hat." }, { "from": "human", "value": "What is the highest accuracy reported in the whole chart?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "Which algorithm has the smallest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Box." }, { "from": "human", "value": "Which algorithm has the largest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Hat." }, { "from": "human", "value": "What is the sum of accuracies of the algorithm hat for all the datasets?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "Is the accuracy of the algorithm hat in the dataset sum smaller than the accuracy of the algorithm box in the dataset access?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What dataset does the steelblue color represent?" }, { "from": "gpt", "value": "Access." }, { "from": "human", "value": "What is the accuracy of the algorithm hat in the dataset sum?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the second group of bars from the left?" }, { "from": "gpt", "value": "Hat." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Access." }, { "from": "human", "value": "How many groups of bars are there?" }, { "from": "gpt", "value": "Two." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099995.png
[ { "from": "human", "value": "How many algorithms have accuracy higher than 20 in at least one dataset?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which algorithm has lowest accuracy for any dataset?" }, { "from": "gpt", "value": "Regard." }, { "from": "human", "value": "What is the lowest accuracy reported in the whole chart?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Which algorithm has the smallest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Regard." }, { "from": "human", "value": "Which algorithm has the largest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Peace." }, { "from": "human", "value": "Is the accuracy of the algorithm regard in the dataset mile smaller than the accuracy of the algorithm peace in the dataset way?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What dataset does the olivedrab color represent?" }, { "from": "gpt", "value": "Way." }, { "from": "human", "value": "What is the accuracy of the algorithm peace in the dataset combat?" }, { "from": "gpt", "value": "70." }, { "from": "human", "value": "What is the label of the first group of bars from the left?" }, { "from": "gpt", "value": "Regard." }, { "from": "human", "value": "What is the label of the third bar from the left in each group?" }, { "from": "gpt", "value": "Way." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00099999.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than 30?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Five." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Bit." }, { "from": "human", "value": "Is the value of genius in phase smaller than the value of item in gold?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the darkorange color represent?" }, { "from": "gpt", "value": "Fall." }, { "from": "human", "value": "What is the value of phase in paint?" }, { "from": "gpt", "value": "80." }, { "from": "human", "value": "What is the label of the fifth group of bars from the left?" }, { "from": "gpt", "value": "Paint." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Phase." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many groups of bars are there?" }, { "from": "gpt", "value": "Five." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100000.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Genius." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Factor." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "60." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "0." }, { "from": "human", "value": "How many algorithms have accuracies lower than 60?" }, { "from": "gpt", "value": "Three." }, { "from": "human", "value": "Is the accuracy of the algorithm march larger than genius?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm factor?" }, { "from": "gpt", "value": "0." }, { "from": "human", "value": "What is the label of the second bar from the left?" }, { "from": "gpt", "value": "Factor." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100005.png
[ { "from": "human", "value": "How many objects are preferred by more than 9 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Origin." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Fluid." }, { "from": "human", "value": "How many total people preferred the object state across all the categories?" }, { "from": "gpt", "value": "17." }, { "from": "human", "value": "Is the object state in the category cup preferred by less people than the object fluid in the category source?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What category does the palegoldenrod color represent?" }, { "from": "gpt", "value": "Cup." }, { "from": "human", "value": "How many people prefer the object origin in the category extent?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the third group of bars from the left?" }, { "from": "gpt", "value": "Yard." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Cup." }, { "from": "human", "value": "How many groups of bars are there?" }, { "from": "gpt", "value": "Seven." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100007.png
[ { "from": "human", "value": "How many objects are preferred by less than 1 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Bed." }, { "from": "human", "value": "How many people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Row." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Bed." }, { "from": "human", "value": "How many total people preferred the object time across all the categories?" }, { "from": "gpt", "value": "12." }, { "from": "human", "value": "What category does the darkorange color represent?" }, { "from": "gpt", "value": "Method." }, { "from": "human", "value": "How many people prefer the object time in the category savage?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "What is the label of the third group of bars from the left?" }, { "from": "gpt", "value": "Time." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Savage." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100010.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than 70?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Which group of bars contains the largest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Moon." }, { "from": "human", "value": "What is the value of the largest individual bar in the whole chart?" }, { "from": "gpt", "value": "80." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "Lot." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Moon." }, { "from": "human", "value": "Is the value of lot in act larger than the value of moon in museum?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the limegreen color represent?" }, { "from": "gpt", "value": "Market." }, { "from": "human", "value": "What is the value of market in moon?" }, { "from": "gpt", "value": "50." }, { "from": "human", "value": "What is the label of the first group of bars from the bottom?" }, { "from": "gpt", "value": "Moon." }, { "from": "human", "value": "What is the label of the second bar from the bottom in each group?" }, { "from": "gpt", "value": "Market." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100012.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than 1000?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Love." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "Love." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Pride." }, { "from": "human", "value": "Are the values in the chart presented in a logarithmic scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the steelblue color represent?" }, { "from": "gpt", "value": "Bird." }, { "from": "human", "value": "What is the value of fit in pride?" }, { "from": "gpt", "value": "1000." }, { "from": "human", "value": "What is the label of the second group of bars from the left?" }, { "from": "gpt", "value": "Love." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Bird." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100014.png
[ { "from": "human", "value": "Which bar has the largest value?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Atom." }, { "from": "human", "value": "Which bar has the smallest value?" }, { "from": "gpt", "value": "Pound." }, { "from": "human", "value": "What is the value of the largest bar?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the value of the smallest bar?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the difference between the largest and the smallest value in the chart?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many bars have values smaller than 9?" }, { "from": "gpt", "value": "Three." }, { "from": "human", "value": "What is the sum of the values of letter and art?" }, { "from": "gpt", "value": "12." }, { "from": "human", "value": "Is the value of atom smaller than pound?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the value of pound?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the third bar from the left?" }, { "from": "gpt", "value": "Letter." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100016.png
[ { "from": "human", "value": "How many items sold more than 7 units in at least one store?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which item sold the most units in any shop?" }, { "from": "gpt", "value": "Seed." }, { "from": "human", "value": "Which item sold the least units in any shop?" }, { "from": "gpt", "value": "Food." }, { "from": "human", "value": "How many units did the best selling item sell in the whole chart?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many units did the worst selling item sell in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which item sold the least number of units summed across all the stores?" }, { "from": "gpt", "value": "Food." }, { "from": "human", "value": "Which item sold the most number of units summed across all the stores?" }, { "from": "gpt", "value": "Seed." }, { "from": "human", "value": "How many units of the item north were sold across all the stores?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What store does the mediumturquoise color represent?" }, { "from": "gpt", "value": "Song." }, { "from": "human", "value": "How many units of the item north were sold in the store song?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "What is the label of the first group of bars from the left?" }, { "from": "gpt", "value": "Food." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Song." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100023.png
[ { "from": "human", "value": "Which bar has the largest value?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Gift." }, { "from": "human", "value": "Which bar has the smallest value?" }, { "from": "gpt", "value": "Wall." }, { "from": "human", "value": "What is the value of the largest bar?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the value of the smallest bar?" }, { "from": "gpt", "value": "-5." }, { "from": "human", "value": "How many bars have values smaller than -5?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Is the value of total larger than gift?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the value of wall?" }, { "from": "gpt", "value": "-5." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Gift." }, { "from": "human", "value": "Does the chart contain any negative values?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100024.png
[ { "from": "human", "value": "How many stacks of bars contain at least one element with value greater than 9?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which stack of bars has the smallest summed value?" }, { "from": "gpt", "value": "Cross." }, { "from": "human", "value": "Which stack of bars has the largest summed value?" }, { "from": "gpt", "value": "Realm." }, { "from": "human", "value": "What is the sum of all the values in the realm group?" }, { "from": "gpt", "value": "22." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the forestgreen color represent?" }, { "from": "gpt", "value": "Stream." }, { "from": "human", "value": "What is the value of stream in realm?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the label of the fourth stack of bars from the left?" }, { "from": "gpt", "value": "Joy." }, { "from": "human", "value": "What is the label of the first element from the bottom in each stack of bars?" }, { "from": "gpt", "value": "Wit." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100029.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Cut." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Turn." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "40." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "30." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "How many algorithms have accuracies lower than 30?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Is the accuracy of the algorithm turn smaller than cut?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm turn?" }, { "from": "gpt", "value": "30." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Cut." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100030.png
[ { "from": "human", "value": "How many objects are preferred by less than 9 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Seven." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Farmer." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Farmer." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Sewage." }, { "from": "human", "value": "How many total people preferred the object place across all the categories?" }, { "from": "gpt", "value": "12." }, { "from": "human", "value": "What category does the lightgoldenrodyellow color represent?" }, { "from": "gpt", "value": "Handle." }, { "from": "human", "value": "How many people prefer the object place in the category handle?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the label of the sixth stack of bars from the left?" }, { "from": "gpt", "value": "Woman." }, { "from": "human", "value": "What is the label of the first element from the bottom in each stack of bars?" }, { "from": "gpt", "value": "Handle." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many stacks of bars are there?" }, { "from": "gpt", "value": "Seven." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100047.png
[ { "from": "human", "value": "How many algorithms have accuracy lower than 90 in at least one dataset?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Which algorithm has the smallest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Block." }, { "from": "human", "value": "Which algorithm has the largest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Dust." }, { "from": "human", "value": "Is the accuracy of the algorithm block in the dataset couple smaller than the accuracy of the algorithm cow in the dataset life?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What dataset does the mediumpurple color represent?" }, { "from": "gpt", "value": "Stand." }, { "from": "human", "value": "What is the accuracy of the algorithm dust in the dataset stand?" }, { "from": "gpt", "value": "70." }, { "from": "human", "value": "What is the label of the third group of bars from the left?" }, { "from": "gpt", "value": "Cow." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Couple." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many groups of bars are there?" }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "How many bars are there per group?" }, { "from": "gpt", "value": "Three." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100049.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Use." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Break." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many algorithms have accuracies higher than 7?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms god and beard?" }, { "from": "gpt", "value": "11." }, { "from": "human", "value": "Is the accuracy of the algorithm beard smaller than use?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm river?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the label of the third bar from the left?" }, { "from": "gpt", "value": "God." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100050.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value smaller than 2?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which group of bars contains the largest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Flux." }, { "from": "human", "value": "What is the value of the largest individual bar in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "King." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Second." }, { "from": "human", "value": "What is the sum of all the values in the second group?" }, { "from": "gpt", "value": "14." }, { "from": "human", "value": "Is the value of king in union larger than the value of bank in hurry?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the steelblue color represent?" }, { "from": "gpt", "value": "Union." }, { "from": "human", "value": "What is the value of hurry in bank?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the label of the first group of bars from the left?" }, { "from": "gpt", "value": "King." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Union." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100055.png
[ { "from": "human", "value": "Which bar has the largest value?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Kid." }, { "from": "human", "value": "Which bar has the smallest value?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "What is the value of the largest bar?" }, { "from": "gpt", "value": "90." }, { "from": "human", "value": "What is the value of the smallest bar?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "What is the difference between the largest and the smallest value in the chart?" }, { "from": "gpt", "value": "80." }, { "from": "human", "value": "How many bars have values smaller than 30?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Is the value of curve smaller than kid?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the value of curve?" }, { "from": "gpt", "value": "50." }, { "from": "human", "value": "What is the label of the fourth bar from the left?" }, { "from": "gpt", "value": "Cry." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100057.png
[ { "from": "human", "value": "How many objects are preferred by more than 1 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Five." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Horror." }, { "from": "human", "value": "How many people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Stand." }, { "from": "human", "value": "How many total people preferred the object belt across all the categories?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Is the object horror in the category dance preferred by less people than the object worry in the category host?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the steelblue color represent?" }, { "from": "gpt", "value": "Host." }, { "from": "human", "value": "How many people prefer the object belt in the category dance?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the label of the fifth group of bars from the left?" }, { "from": "gpt", "value": "Horror." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Dance." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100067.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Page." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Ice." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many objects are liked by more than 9 people?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "How many people prefer the objects page or ice?" }, { "from": "gpt", "value": "11." }, { "from": "human", "value": "Is the object page preferred by more people than ice?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many people prefer the object ice?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Ice." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100068.png
[ { "from": "human", "value": "How many objects are preferred by less than 30 percent of people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Goal." }, { "from": "human", "value": "What percentage of people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "80." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "School." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the green color represent?" }, { "from": "gpt", "value": "Model." }, { "from": "human", "value": "What percentage of people prefer the object theory in the category basis?" }, { "from": "gpt", "value": "30." }, { "from": "human", "value": "What is the label of the fourth group of bars from the bottom?" }, { "from": "gpt", "value": "Theory." }, { "from": "human", "value": "What is the label of the second bar from the bottom in each group?" }, { "from": "gpt", "value": "Model." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100087.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Limit." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Movie." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many algorithms have accuracies lower than 2?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms victim and limit?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Is the accuracy of the algorithm movie larger than victim?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the accuracy of the algorithm limit?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Limit." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100090.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Date." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Pound." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "How many algorithms have accuracies lower than 7?" }, { "from": "gpt", "value": "Three." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms date and pound?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Is the accuracy of the algorithm date smaller than vector?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the accuracy of the algorithm vector?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the fourth bar from the left?" }, { "from": "gpt", "value": "Vector." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100095.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Cast." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Bottom." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many algorithms have accuracies lower than 3?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms bottom and range?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "Is the accuracy of the algorithm bottom smaller than cast?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the accuracy of the algorithm cast?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the label of the second bar from the left?" }, { "from": "gpt", "value": "Cast." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100101.png
[ { "from": "human", "value": "How many algorithms have accuracy higher than 8 in at least one dataset?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Which algorithm has highest accuracy for any dataset?" }, { "from": "gpt", "value": "Track." }, { "from": "human", "value": "What is the highest accuracy reported in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Which algorithm has the smallest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Which algorithm has the largest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Track." }, { "from": "human", "value": "What is the sum of accuracies of the algorithm one for all the datasets?" }, { "from": "gpt", "value": "17." }, { "from": "human", "value": "Is the accuracy of the algorithm track in the dataset west larger than the accuracy of the algorithm winter in the dataset action?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What dataset does the steelblue color represent?" }, { "from": "gpt", "value": "West." }, { "from": "human", "value": "What is the accuracy of the algorithm one in the dataset west?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the third group of bars from the left?" }, { "from": "gpt", "value": "Track." }, { "from": "human", "value": "What is the label of the fourth bar from the left in each group?" }, { "from": "gpt", "value": "Period." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many bars are there per group?" }, { "from": "gpt", "value": "Five." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100104.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than 6?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Which group of bars contains the largest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Artist." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Artist." }, { "from": "human", "value": "What is the value of the largest individual bar in the whole chart?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "What is the sum of all the values in the making group?" }, { "from": "gpt", "value": "17." }, { "from": "human", "value": "Is the value of artist in grade larger than the value of making in blind?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the indianred color represent?" }, { "from": "gpt", "value": "Comedy." }, { "from": "human", "value": "What is the value of comedy in artist?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the label of the first group of bars from the bottom?" }, { "from": "gpt", "value": "Artist." }, { "from": "human", "value": "What is the label of the second bar from the bottom in each group?" }, { "from": "gpt", "value": "Grade." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100109.png
[ { "from": "human", "value": "How many objects are preferred by less than 10 percent of people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Wave." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Latter." }, { "from": "human", "value": "Is the value of wave in notice larger than the value of stem in killer?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the royalblue color represent?" }, { "from": "gpt", "value": "Notice." }, { "from": "human", "value": "What percentage of people prefer the object latter in the category notice?" }, { "from": "gpt", "value": "90." }, { "from": "human", "value": "What is the label of the second group of bars from the bottom?" }, { "from": "gpt", "value": "Wave." }, { "from": "human", "value": "What is the label of the first bar from the bottom in each group?" }, { "from": "gpt", "value": "Notice." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100110.png
[ { "from": "human", "value": "How many items sold less than 9 units in at least one store?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Seven." }, { "from": "human", "value": "Which item sold the least units in any shop?" }, { "from": "gpt", "value": "Bare." }, { "from": "human", "value": "How many units did the worst selling item sell in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which item sold the least number of units summed across all the stores?" }, { "from": "gpt", "value": "Bare." }, { "from": "human", "value": "Which item sold the most number of units summed across all the stores?" }, { "from": "gpt", "value": "None." }, { "from": "human", "value": "How many units of the item stage were sold across all the stores?" }, { "from": "gpt", "value": "15." }, { "from": "human", "value": "Did the item stage in the store dark sold larger units than the item valley in the store inside?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What store does the seagreen color represent?" }, { "from": "gpt", "value": "Dark." }, { "from": "human", "value": "How many units of the item valley were sold in the store dark?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the label of the third group of bars from the left?" }, { "from": "gpt", "value": "Mail." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Dark." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100114.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than 9?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Reader." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "0." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "Reader." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Luxury." }, { "from": "human", "value": "What element does the steelblue color represent?" }, { "from": "gpt", "value": "Ring." }, { "from": "human", "value": "What is the value of tour in stairs?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "What is the label of the second group of bars from the left?" }, { "from": "gpt", "value": "Stairs." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Ring." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100116.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Wagon." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Key." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "How many algorithms have accuracies higher than 4?" }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms wagon and key?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Is the accuracy of the algorithm wagon larger than escape?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a logarithmic scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the accuracy of the algorithm seat?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Union." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100121.png
[ { "from": "human", "value": "How many objects are preferred by more than 1 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Seven." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Suit." }, { "from": "human", "value": "How many people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Arm." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Size." }, { "from": "human", "value": "How many total people preferred the object month across all the categories?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Is the object suit in the category model preferred by more people than the object beard in the category crowd?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the darkorange color represent?" }, { "from": "gpt", "value": "Crowd." }, { "from": "human", "value": "How many people prefer the object beard in the category medium?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "What is the label of the sixth stack of bars from the left?" }, { "from": "gpt", "value": "Ride." }, { "from": "human", "value": "What is the label of the first element from the bottom in each stack of bars?" }, { "from": "gpt", "value": "Medium." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100125.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value smaller than 1?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Flow." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "Flow." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Grant." }, { "from": "human", "value": "What is the sum of all the values in the flow group?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "Is the value of grant in disk smaller than the value of flow in motion?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the mediumturquoise color represent?" }, { "from": "gpt", "value": "Disk." }, { "from": "human", "value": "What is the value of disk in flow?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the second group of bars from the left?" }, { "from": "gpt", "value": "Device." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Motion." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100129.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value smaller than 7?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Labor." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Beauty." }, { "from": "human", "value": "What is the sum of all the values in the whisky group?" }, { "from": "gpt", "value": "15." }, { "from": "human", "value": "Is the value of whisky in family larger than the value of beauty in fear?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the forestgreen color represent?" }, { "from": "gpt", "value": "Fear." }, { "from": "human", "value": "What is the value of fear in beauty?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the label of the third group of bars from the left?" }, { "from": "gpt", "value": "Labor." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Family." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100130.png
[ { "from": "human", "value": "How many objects are preferred by more than 2 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Five." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Run." }, { "from": "human", "value": "How many people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Gay." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Arm." }, { "from": "human", "value": "How many total people preferred the object arm across all the categories?" }, { "from": "gpt", "value": "28." }, { "from": "human", "value": "Is the object arm in the category credit preferred by more people than the object run in the category killer?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many people prefer the object chief in the category killer?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the third group of bars from the left?" }, { "from": "gpt", "value": "Run." }, { "from": "human", "value": "What is the label of the fourth bar from the left in each group?" }, { "from": "gpt", "value": "Credit." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many bars are there per group?" }, { "from": "gpt", "value": "Five." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100131.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value smaller than 30?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Which group of bars contains the largest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Room." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Room." }, { "from": "human", "value": "What is the value of the largest individual bar in the whole chart?" }, { "from": "gpt", "value": "80." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "20." }, { "from": "human", "value": "Is the value of mold in faith smaller than the value of noise in branch?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the darkorange color represent?" }, { "from": "gpt", "value": "Branch." }, { "from": "human", "value": "What is the value of faith in anyone?" }, { "from": "gpt", "value": "70." }, { "from": "human", "value": "What is the label of the seventh group of bars from the left?" }, { "from": "gpt", "value": "Level." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Faith." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100134.png
[ { "from": "human", "value": "How many objects are preferred by more than 90 percent of people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "King." }, { "from": "human", "value": "What percentage of people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "90." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Dear." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "King." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the darkorange color represent?" }, { "from": "gpt", "value": "Mess." }, { "from": "human", "value": "What percentage of people prefer the object dear in the category mess?" }, { "from": "gpt", "value": "70." }, { "from": "human", "value": "What is the label of the second group of bars from the bottom?" }, { "from": "gpt", "value": "Step." }, { "from": "human", "value": "What is the label of the second bar from the bottom in each group?" }, { "from": "gpt", "value": "Mess." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100137.png
[ { "from": "human", "value": "Which bar has the largest value?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Female." }, { "from": "human", "value": "Which bar has the smallest value?" }, { "from": "gpt", "value": "Answer." }, { "from": "human", "value": "What is the value of the largest bar?" }, { "from": "gpt", "value": "90." }, { "from": "human", "value": "What is the value of the smallest bar?" }, { "from": "gpt", "value": "50." }, { "from": "human", "value": "What is the difference between the largest and the smallest value in the chart?" }, { "from": "gpt", "value": "40." }, { "from": "human", "value": "How many bars have values smaller than 60?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Is the value of right smaller than strain?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the value of answer?" }, { "from": "gpt", "value": "50." }, { "from": "human", "value": "What is the label of the seventh bar from the left?" }, { "from": "gpt", "value": "Force." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100146.png
[ { "from": "human", "value": "How many items sold less than 6 units in at least one store?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Which item sold the least units in any shop?" }, { "from": "gpt", "value": "Pure." }, { "from": "human", "value": "How many units did the worst selling item sell in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which item sold the least number of units summed across all the stores?" }, { "from": "gpt", "value": "Cow." }, { "from": "human", "value": "Which item sold the most number of units summed across all the stores?" }, { "from": "gpt", "value": "Dollar." }, { "from": "human", "value": "How many units of the item strike were sold across all the stores?" }, { "from": "gpt", "value": "22." }, { "from": "human", "value": "Did the item strike in the store chief sold larger units than the item cow in the store ladder?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What store does the green color represent?" }, { "from": "gpt", "value": "Ladder." }, { "from": "human", "value": "How many units of the item cow were sold in the store ladder?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the second stack of bars from the left?" }, { "from": "gpt", "value": "Pure." }, { "from": "human", "value": "What is the label of the fourth element from the bottom in each stack of bars?" }, { "from": "gpt", "value": "Ladder." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100149.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Site." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Speech." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many objects are liked by less than 3 people?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "How many people prefer the objects speech or worry?" }, { "from": "gpt", "value": "11." }, { "from": "human", "value": "Is the object speech preferred by more people than worry?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many people prefer the object site?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the label of the second bar from the left?" }, { "from": "gpt", "value": "Father." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Four." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100157.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Party." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Stay." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "How many algorithms have accuracies lower than 6?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms party and stay?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Is the accuracy of the algorithm party smaller than stay?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the accuracy of the algorithm party?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the first bar from the bottom?" }, { "from": "gpt", "value": "Party." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100160.png
[ { "from": "human", "value": "How many objects are preferred by less than 2 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Atom." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Row." }, { "from": "human", "value": "How many people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Row." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Atom." }, { "from": "human", "value": "How many total people preferred the object row across all the categories?" }, { "from": "gpt", "value": "12." }, { "from": "human", "value": "Is the object row in the category method preferred by less people than the object atom in the category act?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the forestgreen color represent?" }, { "from": "gpt", "value": "Act." }, { "from": "human", "value": "How many people prefer the object atom in the category method?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the label of the first stack of bars from the left?" }, { "from": "gpt", "value": "Atom." }, { "from": "human", "value": "What is the label of the first element from the bottom in each stack of bars?" }, { "from": "gpt", "value": "Regard." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100161.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Stake." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Desire." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many objects are liked by less than 4 people?" }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "How many people prefer the objects stake or family?" }, { "from": "gpt", "value": "13." }, { "from": "human", "value": "Is the object world preferred by less people than action?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many people prefer the object output?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the label of the third bar from the left?" }, { "from": "gpt", "value": "End." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Seven." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100164.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Duty." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Mold." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many algorithms have accuracies higher than 6?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms duty and mold?" }, { "from": "gpt", "value": "11." }, { "from": "human", "value": "Is the accuracy of the algorithm mold smaller than nature?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm mold?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the third bar from the bottom?" }, { "from": "gpt", "value": "Duty." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100180.png
[ { "from": "human", "value": "How many stacks of bars contain at least one element with value smaller than 1?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which stack of bars contains the largest valued individual element in the whole chart?" }, { "from": "gpt", "value": "Threat." }, { "from": "human", "value": "What is the value of the largest individual element in the whole chart?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "Which stack of bars has the smallest summed value?" }, { "from": "gpt", "value": "Act." }, { "from": "human", "value": "What is the sum of all the values in the crime group?" }, { "from": "gpt", "value": "12." }, { "from": "human", "value": "Is the value of crime in phase smaller than the value of act in score?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the darkorange color represent?" }, { "from": "gpt", "value": "Phase." }, { "from": "human", "value": "What is the value of phase in edge?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the fourth stack of bars from the bottom?" }, { "from": "gpt", "value": "Crime." }, { "from": "human", "value": "What is the label of the second element from the left in each stack of bars?" }, { "from": "gpt", "value": "Phase." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100185.png
[ { "from": "human", "value": "How many objects are preferred by more than 6 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Share." }, { "from": "human", "value": "How many people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Bench." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Share." }, { "from": "human", "value": "How many total people preferred the object brain across all the categories?" }, { "from": "gpt", "value": "13." }, { "from": "human", "value": "Is the object mile in the category boss preferred by less people than the object share in the category taxi?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the darkorange color represent?" }, { "from": "gpt", "value": "Boss." }, { "from": "human", "value": "How many people prefer the object bench in the category boss?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "What is the label of the fourth group of bars from the left?" }, { "from": "gpt", "value": "Share." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Taxi." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100187.png
[ { "from": "human", "value": "Which bar has the largest value?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Summer." }, { "from": "human", "value": "Which bar has the smallest value?" }, { "from": "gpt", "value": "Cry." }, { "from": "human", "value": "What is the value of the largest bar?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the value of the smallest bar?" }, { "from": "gpt", "value": "-9." }, { "from": "human", "value": "How many bars have values larger than 1?" }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Is the value of summer smaller than south?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the value of team?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the label of the fifth bar from the left?" }, { "from": "gpt", "value": "South." }, { "from": "human", "value": "Does the chart contain any negative values?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Eight." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100189.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Player." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Sugar." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many algorithms have accuracies higher than 1?" }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms sugar and horse?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "Is the accuracy of the algorithm sugar smaller than player?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm sugar?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Sugar." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100202.png
[ { "from": "human", "value": "How many objects are preferred by more than 2 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Foil." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Source." }, { "from": "human", "value": "How many total people preferred the object event across all the categories?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "Is the object foil in the category ocean preferred by less people than the object campus in the category merit?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the darkorange color represent?" }, { "from": "gpt", "value": "Merit." }, { "from": "human", "value": "How many people prefer the object event in the category merit?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the first group of bars from the bottom?" }, { "from": "gpt", "value": "Source." }, { "from": "human", "value": "What is the label of the first bar from the bottom in each group?" }, { "from": "gpt", "value": "Ocean." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100205.png
[ { "from": "human", "value": "How many algorithms have accuracy higher than 7 in at least one dataset?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Which algorithm has highest accuracy for any dataset?" }, { "from": "gpt", "value": "Drug." }, { "from": "human", "value": "Which algorithm has lowest accuracy for any dataset?" }, { "from": "gpt", "value": "Drug." }, { "from": "human", "value": "What is the highest accuracy reported in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the lowest accuracy reported in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which algorithm has the smallest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Bread." }, { "from": "human", "value": "Which algorithm has the largest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Couple." }, { "from": "human", "value": "What is the sum of accuracies of the algorithm drug for all the datasets?" }, { "from": "gpt", "value": "14." }, { "from": "human", "value": "What dataset does the darkorange color represent?" }, { "from": "gpt", "value": "Frame." }, { "from": "human", "value": "What is the accuracy of the algorithm bread in the dataset brick?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the first group of bars from the left?" }, { "from": "gpt", "value": "Drug." }, { "from": "human", "value": "What is the label of the third bar from the left in each group?" }, { "from": "gpt", "value": "Soap." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100206.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value smaller than 3?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Class." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "0." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "Height." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Porch." }, { "from": "human", "value": "Is the value of pipe in kind larger than the value of class in museum?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the darkorange color represent?" }, { "from": "gpt", "value": "Museum." }, { "from": "human", "value": "What is the value of wave in height?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the second group of bars from the left?" }, { "from": "gpt", "value": "Cause." }, { "from": "human", "value": "What is the label of the third bar from the left in each group?" }, { "from": "gpt", "value": "Fault." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100207.png
[ { "from": "human", "value": "How many objects are preferred by less than 1 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Law." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Law." }, { "from": "human", "value": "How many total people preferred the object law across all the categories?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "Is the object west in the category bone preferred by more people than the object yield in the category loan?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What category does the steelblue color represent?" }, { "from": "gpt", "value": "Bone." }, { "from": "human", "value": "How many people prefer the object yield in the category bone?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the label of the first group of bars from the left?" }, { "from": "gpt", "value": "Yield." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Loan." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100211.png
[ { "from": "human", "value": "How many stacks of bars contain at least one element with value smaller than 3?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Which stack of bars contains the smallest valued individual element in the whole chart?" }, { "from": "gpt", "value": "Office." }, { "from": "human", "value": "What is the value of the smallest individual element in the whole chart?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "Which stack of bars has the largest summed value?" }, { "from": "gpt", "value": "Taste." }, { "from": "human", "value": "What is the sum of all the values in the north group?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Is the value of dollar in change smaller than the value of north in silver?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the green color represent?" }, { "from": "gpt", "value": "Change." }, { "from": "human", "value": "What is the value of change in dollar?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "What is the label of the second stack of bars from the bottom?" }, { "from": "gpt", "value": "Taste." }, { "from": "human", "value": "What is the label of the second element from the left in each stack of bars?" }, { "from": "gpt", "value": "Silver." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100212.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Change." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Beauty." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many algorithms have accuracies lower than 8?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms beauty and change?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Is the accuracy of the algorithm change smaller than beauty?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the accuracy of the algorithm change?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Beauty." }, { "from": "human", "value": "Does the chart contain any negative values?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100219.png
[ { "from": "human", "value": "Which item sold the most units?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Whisky." }, { "from": "human", "value": "Which item sold the least units?" }, { "from": "gpt", "value": "Trail." }, { "from": "human", "value": "How many units of the the most sold item were sold?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many units of the the least sold item were sold?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "How many more of the most sold item were sold compared to the least sold item?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "How many items sold less than 5 units?" }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "How many units of items ring and whisky were sold?" }, { "from": "gpt", "value": "13." }, { "from": "human", "value": "Did the item whisky sold more units than summer?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many units of the item summer were sold?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "What is the label of the fourth bar from the left?" }, { "from": "gpt", "value": "Ring." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Four." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100227.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Drug." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Wagon." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many objects are liked by more than 4 people?" }, { "from": "gpt", "value": "Three." }, { "from": "human", "value": "How many people prefer the objects drug or chest?" }, { "from": "gpt", "value": "14." }, { "from": "human", "value": "Is the object wagon preferred by less people than wind?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many people prefer the object wind?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Soul." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100235.png
[ { "from": "human", "value": "How many objects are preferred by less than 9 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Eye." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Eye." }, { "from": "human", "value": "How many people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Flash." }, { "from": "human", "value": "How many total people preferred the object eye across all the categories?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Is the object phase in the category skin preferred by less people than the object eye in the category seat?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What category does the mediumpurple color represent?" }, { "from": "gpt", "value": "Seat." }, { "from": "human", "value": "How many people prefer the object flash in the category seat?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "What is the label of the first group of bars from the left?" }, { "from": "gpt", "value": "Eye." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Skin." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100249.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than 100?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "Travel." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Skin." }, { "from": "human", "value": "Are the values in the chart presented in a logarithmic scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the darkorange color represent?" }, { "from": "gpt", "value": "Army." }, { "from": "human", "value": "What is the value of army in loan?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "What is the label of the fourth group of bars from the bottom?" }, { "from": "gpt", "value": "Run." }, { "from": "human", "value": "What is the label of the second bar from the bottom in each group?" }, { "from": "gpt", "value": "Army." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100251.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Fiber." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Dark." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "80." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "50." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "30." }, { "from": "human", "value": "How many algorithms have accuracies lower than 50?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Is the accuracy of the algorithm fiber larger than dark?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm dark?" }, { "from": "gpt", "value": "50." }, { "from": "human", "value": "What is the label of the first bar from the bottom?" }, { "from": "gpt", "value": "Fiber." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100266.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Uncle." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "70." }, { "from": "human", "value": "How many algorithms have accuracies lower than 10?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Is the accuracy of the algorithm dive larger than uncle?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm ticket?" }, { "from": "gpt", "value": "60." }, { "from": "human", "value": "What is the label of the seventh bar from the bottom?" }, { "from": "gpt", "value": "Wise." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Seven." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100267.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than -3?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Route." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "-3." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "Route." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Boss." }, { "from": "human", "value": "Is the value of boss in rest smaller than the value of route in forest?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the powderblue color represent?" }, { "from": "gpt", "value": "Rest." }, { "from": "human", "value": "What is the value of rest in boss?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "What is the label of the second group of bars from the left?" }, { "from": "gpt", "value": "Boss." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Rest." }, { "from": "human", "value": "Does the chart contain any negative values?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100268.png
[ { "from": "human", "value": "How many objects are preferred by more than 8 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Slave." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Bottom." }, { "from": "human", "value": "How many total people preferred the object sale across all the categories?" }, { "from": "gpt", "value": "15." }, { "from": "human", "value": "Is the object sale in the category trap preferred by more people than the object slave in the category melody?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the royalblue color represent?" }, { "from": "gpt", "value": "Melody." }, { "from": "human", "value": "How many people prefer the object slave in the category desk?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the third group of bars from the left?" }, { "from": "gpt", "value": "Load." }, { "from": "human", "value": "What is the label of the first bar from the left in each group?" }, { "from": "gpt", "value": "Melody." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100277.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Grade." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Slave." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "How many objects are liked by less than 1 people?" }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "How many people prefer the objects shadow or grade?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "Is the object grade preferred by more people than slave?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many people prefer the object shadow?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the third bar from the bottom?" }, { "from": "gpt", "value": "Grade." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100281.png
[ { "from": "human", "value": "Which bar has the largest value?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Wise." }, { "from": "human", "value": "Which bar has the smallest value?" }, { "from": "gpt", "value": "Look." }, { "from": "human", "value": "What is the value of the largest bar?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the value of the smallest bar?" }, { "from": "gpt", "value": "-5." }, { "from": "human", "value": "How many bars have values smaller than 5?" }, { "from": "gpt", "value": "Three." }, { "from": "human", "value": "Is the value of lack larger than army?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the value of army?" }, { "from": "gpt", "value": "-3." }, { "from": "human", "value": "What is the label of the second bar from the bottom?" }, { "from": "gpt", "value": "Look." }, { "from": "human", "value": "Does the chart contain any negative values?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100294.png
[ { "from": "human", "value": "How many objects are preferred by less than 1 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which object is the least preferred in any category?" }, { "from": "gpt", "value": "Dark." }, { "from": "human", "value": "How many people like the least preferred object in the whole chart?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Ten." }, { "from": "human", "value": "How many total people preferred the object goal across all the categories?" }, { "from": "gpt", "value": "13." }, { "from": "human", "value": "Is the object label in the category tone preferred by less people than the object dark in the category essay?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What category does the seagreen color represent?" }, { "from": "gpt", "value": "Essay." }, { "from": "human", "value": "How many people prefer the object dark in the category tone?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the label of the fourth group of bars from the bottom?" }, { "from": "gpt", "value": "Ten." }, { "from": "human", "value": "What is the label of the first bar from the bottom in each group?" }, { "from": "gpt", "value": "Essay." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100302.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value smaller than 1?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "Notice." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Vein." }, { "from": "human", "value": "Is the value of vein in risk larger than the value of notice in critic?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the olivedrab color represent?" }, { "from": "gpt", "value": "Critic." }, { "from": "human", "value": "What is the value of risk in notice?" }, { "from": "gpt", "value": "-7." }, { "from": "human", "value": "What is the label of the second group of bars from the left?" }, { "from": "gpt", "value": "Union." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Excess." }, { "from": "human", "value": "Does the chart contain any negative values?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100307.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value greater than 1?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Six." }, { "from": "human", "value": "Which group has the smallest summed value?" }, { "from": "gpt", "value": "Cloth." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Charge." }, { "from": "human", "value": "Is the value of reply in engine smaller than the value of route in strain?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What element does the steelblue color represent?" }, { "from": "gpt", "value": "Unit." }, { "from": "human", "value": "What is the value of strain in beach?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the label of the seventh group of bars from the left?" }, { "from": "gpt", "value": "Beach." }, { "from": "human", "value": "What is the label of the third bar from the left in each group?" }, { "from": "gpt", "value": "Engine." }, { "from": "human", "value": "Does the chart contain stacked bars?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many groups of bars are there?" }, { "from": "gpt", "value": "Seven." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100313.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Drop." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Jet." }, { "from": "human", "value": "What percentage of people prefer the most preferred object?" }, { "from": "gpt", "value": "90." }, { "from": "human", "value": "What percentage of people prefer the least preferred object?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "80." }, { "from": "human", "value": "How many objects are liked by more than 20 percent of people?" }, { "from": "gpt", "value": "Five." }, { "from": "human", "value": "Is the object honor preferred by less people than drop?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What percentage of people prefer the object flux?" }, { "from": "gpt", "value": "60." }, { "from": "human", "value": "What is the label of the fourth bar from the left?" }, { "from": "gpt", "value": "Vein." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100322.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Fit." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Cover." }, { "from": "human", "value": "What percentage of people prefer the most preferred object?" }, { "from": "gpt", "value": "80." }, { "from": "human", "value": "What percentage of people prefer the least preferred object?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "70." }, { "from": "human", "value": "How many objects are liked by less than 60 percent of people?" }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Is the object cover preferred by less people than band?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What percentage of people prefer the object clay?" }, { "from": "gpt", "value": "40." }, { "from": "human", "value": "What is the label of the third bar from the left?" }, { "from": "gpt", "value": "Cover." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100325.png
[ { "from": "human", "value": "How many groups of bars contain at least one bar with value smaller than 10?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Zero." }, { "from": "human", "value": "Which group of bars contains the smallest valued individual bar in the whole chart?" }, { "from": "gpt", "value": "Bass." }, { "from": "human", "value": "What is the value of the smallest individual bar in the whole chart?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "Which group has the largest summed value?" }, { "from": "gpt", "value": "Bass." }, { "from": "human", "value": "Is the value of boss in lunch larger than the value of bass in row?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What element does the crimson color represent?" }, { "from": "gpt", "value": "Row." }, { "from": "human", "value": "What is the value of lunch in boss?" }, { "from": "gpt", "value": "20." }, { "from": "human", "value": "What is the label of the second group of bars from the left?" }, { "from": "gpt", "value": "Ship." }, { "from": "human", "value": "What is the label of the third bar from the left in each group?" }, { "from": "gpt", "value": "Lunch." }, { "from": "human", "value": "How many bars are there per group?" }, { "from": "gpt", "value": "Four." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100326.png
[ { "from": "human", "value": "Which item sold the most units?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Salt." }, { "from": "human", "value": "Which item sold the least units?" }, { "from": "gpt", "value": "Device." }, { "from": "human", "value": "How many units of the the most sold item were sold?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "How many units of the the least sold item were sold?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "How many more of the most sold item were sold compared to the least sold item?" }, { "from": "gpt", "value": "7." }, { "from": "human", "value": "How many items sold more than 2 units?" }, { "from": "gpt", "value": "Three." }, { "from": "human", "value": "How many units of items task and safety were sold?" }, { "from": "gpt", "value": "10." }, { "from": "human", "value": "How many units of the item safety were sold?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "What is the label of the fourth bar from the bottom?" }, { "from": "gpt", "value": "Gas." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100328.png
[ { "from": "human", "value": "How many objects are preferred by less than 60 percent of people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Fit." }, { "from": "human", "value": "What percentage of people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "90." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Tire." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Sex." }, { "from": "human", "value": "Is the value of tire in stay larger than the value of sex in neck?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the values in the chart presented in a percentage scale?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What category does the darkturquoise color represent?" }, { "from": "gpt", "value": "Neck." }, { "from": "human", "value": "What percentage of people prefer the object tire in the category neck?" }, { "from": "gpt", "value": "60." }, { "from": "human", "value": "What is the label of the second group of bars from the bottom?" }, { "from": "gpt", "value": "Fit." }, { "from": "human", "value": "What is the label of the second bar from the bottom in each group?" }, { "from": "gpt", "value": "Stay." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100331.png
[ { "from": "human", "value": "How many algorithms have accuracy lower than 6 in at least one dataset?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Four." }, { "from": "human", "value": "Which algorithm has the smallest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Campus." }, { "from": "human", "value": "Which algorithm has the largest accuracy summed across all the datasets?" }, { "from": "gpt", "value": "Study." }, { "from": "human", "value": "What is the sum of accuracies of the algorithm step for all the datasets?" }, { "from": "gpt", "value": "24." }, { "from": "human", "value": "Is the accuracy of the algorithm step in the dataset sum smaller than the accuracy of the algorithm campus in the dataset style?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What dataset does the darkred color represent?" }, { "from": "gpt", "value": "Style." }, { "from": "human", "value": "What is the accuracy of the algorithm step in the dataset style?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the label of the third group of bars from the left?" }, { "from": "gpt", "value": "Step." }, { "from": "human", "value": "What is the label of the third bar from the left in each group?" }, { "from": "gpt", "value": "Style." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100343.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Market." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Pool." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "5." }, { "from": "human", "value": "How many algorithms have accuracies lower than 8?" }, { "from": "gpt", "value": "Five." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms order and pool?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "Are the values in the chart presented in a logarithmic scale?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the accuracy of the algorithm stick?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the label of the sixth bar from the left?" }, { "from": "gpt", "value": "Weight." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Seven." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100346.png
[ { "from": "human", "value": "Which object is the most preferred?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Lack." }, { "from": "human", "value": "Which object is the least preferred?" }, { "from": "gpt", "value": "Judge." }, { "from": "human", "value": "How many people prefer the most preferred object?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "How many people prefer the least preferred object?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the difference between most and least preferred object?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many objects are liked by less than 8 people?" }, { "from": "gpt", "value": "Two." }, { "from": "human", "value": "How many people prefer the objects pair or lack?" }, { "from": "gpt", "value": "14." }, { "from": "human", "value": "Is the object lack preferred by more people than judge?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "How many people prefer the object lack?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "What is the label of the second bar from the bottom?" }, { "from": "gpt", "value": "Pair." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "Yes." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100347.png
[ { "from": "human", "value": "How many objects are preferred by more than 7 people in at least one category?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "One." }, { "from": "human", "value": "Which object is the most preferred in any category?" }, { "from": "gpt", "value": "Crime." }, { "from": "human", "value": "How many people like the most preferred object in the whole chart?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "Which object is preferred by the least number of people summed across all the categories?" }, { "from": "gpt", "value": "Anyone." }, { "from": "human", "value": "Which object is preferred by the most number of people summed across all the categories?" }, { "from": "gpt", "value": "Crime." }, { "from": "human", "value": "How many total people preferred the object brief across all the categories?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "Is the object anyone in the category coffee preferred by more people than the object rock in the category pipe?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What category does the lime color represent?" }, { "from": "gpt", "value": "Coffee." }, { "from": "human", "value": "How many people prefer the object rock in the category coffee?" }, { "from": "gpt", "value": "2." }, { "from": "human", "value": "What is the label of the second group of bars from the left?" }, { "from": "gpt", "value": "Butter." }, { "from": "human", "value": "What is the label of the second bar from the left in each group?" }, { "from": "gpt", "value": "Coffee." }, { "from": "human", "value": "Is each bar a single solid color without patterns?" }, { "from": "gpt", "value": "No." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100349.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Share." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Color." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "6." }, { "from": "human", "value": "How many algorithms have accuracies higher than 4?" }, { "from": "gpt", "value": "Five." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms bay and share?" }, { "from": "gpt", "value": "15." }, { "from": "human", "value": "Is the accuracy of the algorithm trust smaller than news?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "What is the accuracy of the algorithm news?" }, { "from": "gpt", "value": "4." }, { "from": "human", "value": "What is the label of the third bar from the left?" }, { "from": "gpt", "value": "Bay." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Eight." } ]
dvqa(cauldron,llava_format)
cauldron/dvqa/images/dvqa_00100350.png
[ { "from": "human", "value": "Which algorithm has the highest accuracy?\nAnswer the question using a single word or phrase." }, { "from": "gpt", "value": "Grain." }, { "from": "human", "value": "Which algorithm has the lowest accuracy?" }, { "from": "gpt", "value": "Title." }, { "from": "human", "value": "What is the accuracy of the algorithm with highest accuracy?" }, { "from": "gpt", "value": "9." }, { "from": "human", "value": "What is the accuracy of the algorithm with lowest accuracy?" }, { "from": "gpt", "value": "1." }, { "from": "human", "value": "How much more accurate is the most accurate algorithm compared the least accurate algorithm?" }, { "from": "gpt", "value": "8." }, { "from": "human", "value": "How many algorithms have accuracies lower than 6?" }, { "from": "gpt", "value": "Three." }, { "from": "human", "value": "What is the sum of the accuracies of the algorithms crime and grain?" }, { "from": "gpt", "value": "16." }, { "from": "human", "value": "Is the accuracy of the algorithm team smaller than crime?" }, { "from": "gpt", "value": "Yes." }, { "from": "human", "value": "What is the accuracy of the algorithm judge?" }, { "from": "gpt", "value": "3." }, { "from": "human", "value": "What is the label of the first bar from the left?" }, { "from": "gpt", "value": "Judge." }, { "from": "human", "value": "Does the chart contain any negative values?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "Are the bars horizontal?" }, { "from": "gpt", "value": "No." }, { "from": "human", "value": "How many bars are there?" }, { "from": "gpt", "value": "Six." } ]
dvqa(cauldron,llava_format)