Scientific Reasoning on GPQA Diamond (avg@3)
82.2avg@3Nanbeige4-3B-Thinking
Evaluation Results
| Method | Links | |
|---|---|---|
| Nanbeige4-3B-ThinkingRelease Identifier=25112025.12 | 82.2 | |
| Qwen3-30A3-2507Release Identifier=25072025.12 | 73.4 | |
| Qwen3-32B-2504Release Identifier=25042025.12 | 68.7 | |
| Qwen3-4B-2507Release Identifier=25072025.12 | 67.2 | |
| Qwen3-14B-2504Release Identifier=25042025.12 | 64 | |
| Qwen3-8B-2504Release Identifier=25042025.12 | 62 |