List your company

Error Detection on 74-paper benchmark (full)

55.4RecallOpenAIReview
10.36822.05933.7545.441Jun 18, 2026

Evaluation results

MethodLinks
OpenAIReviewBackend=DeepSeek-V4-Flash2026.0655.4
OpenAIReviewBackend=Grok-4.1-Fast2026.0649.3
OpenAIReviewBackend=Qwen3.6-35B-A3B2026.0646.2
zero-shotBackend=Grok-4.1-Fast2026.0633.5
zero-shotBackend=Qwen3.6-35B-A3B2026.0631.7
OpenAIReviewBackend=Gemini-3.1-Flash-Lite2026.0630.9
zero-shotBackend=DeepSeek-V4-Flash2026.0629.4
coarseBackend=DeepSeek-V4-Flash2026.0621.2
coarseBackend=Qwen3.6-35B-A3B2026.0615.9
coarseBackend=Grok-4.1-Fast2026.0614.8
zero-shotBackend=Gemini-3.1-Flash-Lite2026.0614.3
coarseBackend=Gemini-3.1-Flash-Lite2026.0612.1