Natural Language Visual Reasoning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
92.6Accuracy
346
Mar 18, 2026
91.51Accuracy
307
Mar 18, 2026
88.8Accuracy
41
Apr 21, 2026
85.36Accuracy
16
Feb 26, 2026
85.5Accuracy
14
Feb 26, 2026
83.15Accuracy
12
Feb 26, 2026
17.4GFLOPs
4
Feb 26, 2026
67.3Accuracy
2
Feb 26, 2026