Visual Question Answering
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
43.5Accuracy
2
Feb 26, 2026
94.8Accuracy
2
Feb 26, 2026
70.2Accuracy
2
Feb 26, 2026
56.4Exact-match Accuracy (en)
2
Feb 26, 2026
73.7Accuracy
2
Feb 26, 2026
99.96Systematic Generalization Accuracy
2
Feb 27, 2026
55.1Accuracy (All)
2
Feb 26, 2026
87.91QAAS (%)
2
Feb 26, 2026
24BLEU
1
May 26, 2026
83.99Accuracy
1
Feb 26, 2026
62.5Baseline Performance
1
Feb 26, 2026
78.2Baseline Perf.
1
Feb 26, 2026
—
—Primary metric
0
May 26, 2026
—
—Primary metric
0
May 26, 2026
—
—Accuracy
0
Feb 26, 2026
—
—Accuracy
0
Feb 26, 2026
—
—Presence Accuracy
0
Feb 26, 2026
—
—Accuracy
0
Feb 26, 2026
—
—Accuracy
0
Feb 26, 2026
—
—Accuracy
0
Feb 26, 2026
—
—Accuracy
0
Feb 26, 2026
—
—Accuracy
0
Feb 26, 2026