Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Cross-distribution performance on SAIR (hard2)

99Accuracy

98_hard200

38.6854.347085.66Apr 20, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.04
99
2026.04
97
2026.04
89
2026.04
86
2026.04
83
2026.04
80
2026.04
60
2026.04
58
2026.04
51.5
2026.04
51
2026.04
49.5
2026.04
48
2026.04
41