Faithfulness detection on ProcessBench
83.2F1 ScoreGPT-o1
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-o12026.05 | 83.2 | |
| GPT-4o2026.05 | 82.9 | |
| o3-mini2026.05 | 73.8 | |
| GeoFaith2026.05 | 65.7 | |
| DeepSeek-V32026.05 | 63.5 | |
| LogicReward2026.05 | 53.8 | |
| Llama-3.1-70B-Instruct2026.05 | 49.8 | |
| Qwen2.5-32B-Instruct2026.05 | 47.3 | |
| FaithLens2026.05 | 23.2 | |
| HHEM2.12026.05 | 17.9 |