Causal Reasoning on Cladder (Accuracy)
82.7AccuracyCDCR
Evaluation Results
| Method | Links | |
|---|---|---|
| CDCRBase Model=Qwen3-8B, Data=CDCR2026.05 | 82.7 | |
| Graph-Guided (Ours)2026.05 | 82.3 | |
| Qwen3-32BBase Model=Qwen3-32B2026.05 | 82.3 | |
| CDCRBase Model=Qwen3-4B, Data=CDCR2026.05 | 82 | |
| CDCRBase Model=Olmo3-7B-Instruct, Data=CDCR2026.05 | 81.7 | |
| UNICOBase Model=Qwen3-8B, Data=UNICO2026.05 | 81 | |
| UNICOBase Model=Qwen3-4B, Data=UNICO2026.05 | 79 | |
| Olmo3.1-32B-InstructBase Model=Olmo3.1-32B-Instruct2026.05 | 78 | |
| UNICOBase Model=Olmo3-7B-Instruct, Data=UNICO2026.05 | 77.3 | |
| CauGymBase Model=Qwen3-8B, Data=CauGym2026.05 | 76.7 | |
| CauGymBase Model=Qwen3-4B, Data=CauGym2026.05 | 75.2 | |
| OriginalBase Model=Qwen3-8B, Data=Original2026.05 | 74.9 | |
| CauGymBase Model=Olmo3-7B-Instruct, Data=CauGym2026.05 | 74.3 | |
| OriginalBase Model=Olmo3-7B-Instruct, Data=Original2026.05 | 74.2 | |
| OriginalBase Model=Qwen3-4B, Data=Original2026.05 | 71.2 | |
| Causal CoT2026.05 | 70.4 | |
| GPT-42026.05 | 62 | |
| GPT-3.52026.05 | 52.2 | |
| Alpaca2026.05 | 44.7 | |
| LLaMa2026.05 | 44 |