Procedural Reasoning on RecipeQA
46.6BertScoreGPT-4o
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT-4oFramework=CoP2026.05 | 46.6 | 49.7 | |
| Gemini-2.5Framework=Baseline2026.05 | 42.8 | 47.2 | |
| GPT-4oFramework=Baseline2026.05 | 38.7 | 48.6 | |
| Gemini-2.5Framework=CoP2026.05 | 37.8 | 44.6 | |
| Qwen2.5-72BFramework=CoP2026.05 | 26.5 | 32.3 | |
| Qwen2.5-72BFramework=Baseline2026.05 | 23.9 | 35 | |
| GPT-4o-miniFramework=CoP2026.05 | 22 | 29.6 | |
| GPT-4o-miniFramework=Baseline2026.05 | 20.7 | 31.9 | |
| Claude-3-5Framework=CoP2026.05 | 19.3 | 33.6 | |
| Claude-3-5Framework=Baseline2026.05 | 16.8 | 26.3 |