Inductive Reasoning on Clutrr (Pass@1)
95.5Pass@1RULEREASONER-8B
Evaluation results
| Method | Links | ||
|---|---|---|---|
| RULEREASONER-8BMethod Category=RULEREASONER (Ours), Model Scale=8B2025.06 | 95.5 | 0.3 | |
| ADARFTMethod Category=CURRICULUM LEARNING2025.06 | 92.5 | — | |
| Easy-to-hard RLMethod Category=CURRICULUM LEARNING2025.06 | 88 | — | |
| DAPOMethod Category=ADVANCED RLVRS2025.06 | 86.5 | — | |
| Data-balance RLMethod Category=CURRICULUM LEARNING2025.06 | 86.5 | — | |
| SFT w/ Long CoTMethod Category=BEHAVIORAL CLONING, Reasoning Strategy=Long CoT2025.06 | 83.5 | — | |
| RULEREASONER-4BMethod Category=RULEREASONER (Ours), Model Scale=4B2025.06 | 82 | 0.4 | |
| SFT w/ Short CoTMethod Category=BEHAVIORAL CLONING, Reasoning Strategy=Short CoT2025.06 | 77.6 | — | |
| GRPOMethod Category=ADVANCED RLVRS2025.06 | 73.1 | — | |
| DeepSeek-R1Method Category=FRONTIER REASONERS2025.06 | 71.6 | — | |
| Dr. GRPOMethod Category=ADVANCED RLVRS2025.06 | 68.6 | — | |
| Claude-3.7-SonnetMethod Category=FRONTIER REASONERS2025.06 | 65.7 | — | |
| OpenAI o1Method Category=FRONTIER REASONERS2025.06 | 52.2 | — | |
| Chain-of-LogicMethod Category=PRIOR RBRS2025.06 | 44.8 | — | |
| HtTMethod Category=PRIOR RBRS2025.06 | 40.3 | — | |
| OpenAI o3-miniMethod Category=FRONTIER REASONERS2025.06 | 40.3 | — | |
| SFT w/o CoTMethod Category=BEHAVIORAL CLONING, Reasoning Strategy=w/o CoT2025.06 | 37.5 | — | |
| RGFBMethod Category=PRIOR RBRS2025.06 | 31.3 | — |