Logical Reasoning on Web of Lies
59.2PerformanceDART
Evaluation Results
| Method | Links | |
|---|---|---|
| DARTBase Model=Qwen3-0.6B2026.05 | 59.2 | |
| STaRBase Model=Qwen3-0.6B2026.05 | 53.4 | |
| TESSYBase Model=Qwen3-0.6B2026.05 | 52.4 | |
| BaseBase Model=Qwen3-0.6B2026.05 | 50.5 | |
| GRPOBase Model=Qwen3-0.6B2026.05 | 49.5 | |
| DARTBase Model=Qwen2.5-0.5B-Instruct2026.05 | 46.9 | |
| BaseBase Model=Qwen2.5-0.5B-Instruct2026.05 | 41.1 | |
| STaRBase Model=Qwen2.5-0.5B-Instruct2026.05 | 40.4 | |
| GRPOBase Model=Qwen2.5-0.5B-Instruct2026.05 | 39.7 | |
| Original-SFTBase Model=Qwen2.5-0.5B-Instruct2026.05 | 29.8 | |
| TESSYBase Model=Qwen2.5-0.5B-Instruct2026.05 | 25.8 | |
| Original-SFTBase Model=Qwen3-0.6B2026.05 | 20.2 |