List your company

General Reasoning on GPQA Diamond (Score)

85.4General Reasoning ScoreConfident Decoding
57.00864.37971.7579.121Jun 20, 2026

Evaluation results

MethodLinks
Confident DecodingModel=Qwen3.5-122B-A10B, Decoding Strategy=Confident (Ours)2026.0685.4
Last LayerModel=Qwen3.5-122B-A10B, Decoding Strategy=Last Layer2026.0683.3
Confident DecodingModel=Qwen3.5-35B-A3B, Decoding Strategy=Confident (Ours)2026.0682.8
Confident DecodingModel=Gemma-4-31B, Decoding Strategy=Confident (Ours)2026.0680.8
Confident DecodingModel=gpt-oss-120B, Decoding Strategy=Confident (Ours)2026.0680.8
Confident DecodingModel=Qwen3.5-27B, Decoding Strategy=Confident (Ours)2026.0679.4
Last LayerModel=Qwen3.5-27B, Decoding Strategy=Last Layer2026.0678.2
Last LayerModel=Gemma-4-31B, Decoding Strategy=Last Layer2026.0676.8
Last LayerModel=Qwen3.5-35B-A3B, Decoding Strategy=Last Layer2026.0676.3
Last LayerModel=gpt-oss-120B, Decoding Strategy=Last Layer2026.0676.3
Confident DecodingModel=gpt-oss-20b, Decoding Strategy=Confident (Ours)2026.0660.8
Last LayerModel=gpt-oss-20b, Decoding Strategy=Last Layer2026.0658.1