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

Error Detection on Perturbation Benchmark 24-paper frontier

0.716Recall

OpenAIReview

0.099280.259390.41950.57961Jun 18, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
0.716
2026.06
0.681
2026.06
0.598
2026.06
0.558
2026.06
0.543
2026.06
0.526
2026.06
0.454
2026.06
0.33
2026.06
0.314
2026.06
0.314
2026.06
0.31
0.265
2026.06
0.207
2026.06
0.182
2026.06
0.147
2026.06
0.128
2026.06
0.123