Code Reasoning on LCB
62.46pass@1SCF-RKL
Evaluation Results
| Method | Links | |
|---|---|---|
| SCF-RKLModels=Fuse2026.02 | 62.46 | |
| w/o MergingModels=Math2026.02 | 55.91 | |
| w/o MergingModels=Code2026.02 | 55.2 | |
| SCEModels=Fuse2026.02 | 44.35 | |
| Ties MergingModels=Fuse2026.02 | 44.18 | |
| Dare Ties MergingModels=Fuse2026.02 | 40.05 | |
| Task ArithmeticModels=Fuse2026.02 | 38.53 | |
| Dare Task ArithmeticModels=Fuse2026.02 | 38.53 | |
| Teacher-OPPOBase Model=Qwen3-4B-Instruct-2507, Optimization Framework=DAPO2026.05 | 36.2 | |
| Teacher-OPPOBase Model=Qwen3-4B-Instruct-2507, Optimization Framework=GRPO2026.05 | 35.4 | |
| Self-OPPOBase Model=Qwen3-4B-Instruct-2507, Optimization Framework=DAPO2026.05 | 34.6 | |
| Self-OPPOBase Model=Qwen3-4B-Instruct-2507, Optimization Framework=GRPO2026.05 | 34.2 | |
| DAPOBase Model=Qwen3-4B-Instruct-25072026.05 | 32.8 | |
| SDPOBase Model=Qwen3-4B-Instruct-25072026.05 | 32.4 | |
| GRPOBase Model=Qwen3-4B-Instruct-25072026.05 | 31.5 | |
| Dr.GRPOBase Model=Qwen3-4B-Instruct-25072026.05 | 31.4 | |
| SFTBase Model=Qwen3-4B-Instruct-25072026.05 | 28.6 | |
| Teacher-OPPOBase Model=Phi-4-mini-instruct, Optimization Framework=DAPO2026.05 | 28 | |
| Teacher-OPPOBase Model=Phi-4-mini-instruct, Optimization Framework=GRPO2026.05 | 27.2 | |
| Self-OPPOBase Model=Phi-4-mini-instruct, Optimization Framework=DAPO2026.05 | 26.5 | |
| Self-OPPOBase Model=Phi-4-mini-instruct, Optimization Framework=GRPO2026.05 | 25.8 | |
| DAPOBase Model=Phi-4-mini-instruct2026.05 | 24.2 | |
| SDPOBase Model=Phi-4-mini-instruct2026.05 | 23.8 | |
| GRPOBase Model=Phi-4-mini-instruct2026.05 | 23 | |
| Dr.GRPOBase Model=Phi-4-mini-instruct2026.05 | 22.8 | |
| SFTBase Model=Phi-4-mini-instruct2026.05 | 20.4 |