Graduate-level Reasoning on GPQA diamond (Pass@10)
78.29Pass@10Multi-chain
Evaluation Results
| Method | Links | |
|---|---|---|
| Multi-chainModel=Qwen2.5-32B-Instruct2026.01 | 78.29 | |
| Entropy-TreeModel=Qwen2.5-32B-Instruct2026.01 | 77.98 | |
| Entropy-TreeModel=Qwen2.5-7B-Instruct2026.01 | 74.81 | |
| Entropy-TreeModel=Qwen2.5-14B-Instruct2026.01 | 74.1 | |
| Multi-chainModel=Qwen2.5-14B-Instruct2026.01 | 73.25 | |
| Multi-chainModel=Qwen2.5-7B-Instruct2026.01 | 70.83 |