Logical Reasoning on ProverQA hard split
0.686AccuracyLogicAgent
Evaluation Results
| Method | Links | |
|---|---|---|
| LogicAgentType=AR+SR, Base Model=Qwen2.52025.09 | 0.686 | |
| Logic-LMType=SR, Base Model=Qwen2.52025.09 | 0.624 | |
| SymbCoTType=SR, Base Model=Qwen2.52025.09 | 0.572 | |
| AristotleType=AR+SR, Base Model=Qwen2.52025.09 | 0.562 | |
| ToTType=AR, Base Model=Qwen2.52025.09 | 0.534 | |
| CRType=AR, Base Model=Qwen2.52025.09 | 0.518 | |
| CoTType=LR, Base Model=Qwen2.52025.09 | 0.472 | |
| NaiveType=LR, Base Model=Qwen2.52025.09 | 0.396 |