General Knowledge on GPQA Diamond
65Pass@1Claude 3.5 Sonnet-1022
Evaluation results
| Method | Links | |
|---|---|---|
| Claude 3.5 Sonnet-1022Model Backbone=Claude 3.5 Sonnet2026.03 | 65 | |
| MiMo-RL 7B-R-TAPModel Backbone=7B, Reasoning Strategy=MiMo-RL + R-TAP2026.03 | 60.7 | |
| OpenAI o1-miniModel Backbone=o1-mini2026.03 | 60 | |
| R1-Distill-Qwen-14BModel Backbone=Qwen-14B, Reasoning Strategy=R1-Distill2026.03 | 59.1 | |
| QwQ-32B PreviewModel Backbone=QwQ-32B2026.03 | 54.5 | |
| MiMo 7B-RLModel Backbone=7B, Reasoning Strategy=MiMo-RL2026.03 | 54.4 | |
| GPT-4o 0513Model Backbone=GPT-4o2026.03 | 49.9 | |
| R1-Distill-Qwen-7BModel Backbone=Qwen-7B, Reasoning Strategy=R1-Distill2026.03 | 49.1 | |
| SRFT [13]Base Model=Qwen2.5-Math-7B, Source=reported2026.04 | 46.4 | |
| ReLIFT [12]Base Model=Qwen2.5-Math-7B, Source=reported2026.04 | 43.1 | |
| HPT [15]Base Model=Qwen2.5-Math-7B, Source=reported2026.04 | 42.9 | |
| LUFFY [11]Base Model=Qwen2.5-Math-7B, Source=reproduction attempt2026.04 | 41.8 | |
| SFT→RLBase Model=Qwen2.5-Math-7B, Source=verl2026.04 | 40.6 | |
| LUFFY [11]Base Model=Qwen2.5-Math-7B, Source=reported2026.04 | 39.9 | |
| Prefix-RFT [14]Base Model=Qwen2.5-Math-7B, Source=reported2026.04 | 39.1 | |
| ReLIFT [12]Base Model=Qwen2.5-Math-7B, Source=reproduction attempt2026.04 | 36.7 | |
| SFTBase Model=Qwen2.5-Math-7B, Source=verl2026.04 | 24.6 |