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

Error Detection on 74-paper benchmark (full)

55.4Recall

OpenAIReview

10.36822.05933.7545.441Jun 18, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
55.4
2026.06
49.3
2026.06
46.2
2026.06
33.5
2026.06
31.7
2026.06
30.9
2026.06
29.4
2026.06
21.2
2026.06
15.9
2026.06
14.8
2026.06
14.3
2026.06
12.1