Reasoning on PhantomWiki
51.2ScoreDOLORES
Evaluation Results
| Method | Links | |
|---|---|---|
| DOLORESModel=Llama-3.3 70B Instruct2026.05 | 51.2 | |
| CodeActModel=Llama-3.3 70B Instruct2026.05 | 38.1 | |
| DOLORESModel=Qwen3 32B Thinking2026.05 | 36.9 | |
| RLMModel=Llama-3.3 70B Instruct2026.05 | 27 | |
| CodeActModel=Qwen3 32B Thinking2026.05 | 25.2 | |
| ReActModel=Llama-3.3 70B Instruct2026.05 | 24.4 | |
| Deep ResearchModel=Qwen3 32B Thinking2026.05 | 21.2 | |
| DOLORESModel=Qwen3 8B Thinking2026.05 | 17.2 | |
| ReActModel=Qwen3 32B Thinking2026.05 | 16.7 | |
| RLMModel=Qwen3 32B Thinking2026.05 | 16 | |
| Deep ResearchModel=Llama-3.3 70B Instruct2026.05 | 16 | |
| CodeActModel=Qwen3 8B Thinking2026.05 | 15.3 | |
| ReActModel=Qwen3 8B Thinking2026.05 | 14.4 | |
| Deep ResearchModel=Qwen3 8B Thinking2026.05 | 12 | |
| RLMModel=Qwen3 8B Thinking2026.05 | 4.3 |