Shared
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Shared (evaluation)
78Tie-aware Accuracy
10
Shared 500-prompt sample
0.87Global Correlation (r)
5
Shared 40-task (evaluation)
40Win Count
4
Shared 20-object (test)
74.7Affordance Realism
4
Shared pooled aggregation (test)
0.1Brier Score (BS)
4