General Knowledge Reasoning on Humanity's Last Exam
50ScoreFugu-Ultra
Evaluation Results
| Method | Links | |
|---|---|---|
| Fugu-Ultra2026.06 | 50 | |
| Claude Opus 4.82026.06 | 49.8 | |
| Fugu2026.06 | 47.2 | |
| Gemini 3.12026.06 | 44.4 | |
| GPT-5.52026.06 | 41.4 | |
| Confident DecodingModel=Qwen3.5-27B, Decoding Strategy=Confident (Ours)2026.06 | 16.8 | |
| Confident DecodingModel=Qwen3.5-122B-A10B, Decoding Strategy=Confident (Ours)2026.06 | 16.5 | |
| Last LayerModel=Qwen3.5-27B, Decoding Strategy=Last Layer2026.06 | 16 | |
| Last LayerModel=Qwen3.5-122B-A10B, Decoding Strategy=Last Layer2026.06 | 14.7 | |
| Confident DecodingModel=gpt-oss-120B, Decoding Strategy=Confident (Ours)2026.06 | 12.6 | |
| Confident DecodingModel=Qwen3.5-35B-A3B, Decoding Strategy=Confident (Ours)2026.06 | 11.2 | |
| Confident DecodingModel=Gemma-4-31B, Decoding Strategy=Confident (Ours)2026.06 | 11 | |
| Last LayerModel=gpt-oss-120B, Decoding Strategy=Last Layer2026.06 | 10.8 | |
| Last LayerModel=Qwen3.5-35B-A3B, Decoding Strategy=Last Layer2026.06 | 9.2 | |
| Last LayerModel=Gemma-4-31B, Decoding Strategy=Last Layer2026.06 | 9.2 | |
| Confident DecodingModel=gpt-oss-20b, Decoding Strategy=Confident (Ours)2026.06 | 6.3 | |
| Last LayerModel=gpt-oss-20b, Decoding Strategy=Last Layer2026.06 | 6.2 |