Multiple-choice Question Answering on ARC-Easy zero-shot (test)
94.2AccuracyRLVR
Evaluation results
| Method | Links | |
|---|---|---|
| RLVRBackbone=Qwen2.5-7B-Instruct, Training=RLVR, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 94.2 | |
| Qwen2.5-7B-InstructBackbone=Qwen2.5-7B-Instruct, Few-shot=zero-shot2026.03 | 93.8 | |
| MIPOBackbone=Qwen2.5-7B-Instruct, Training=MIPO, Few-shot=zero-shot2026.03 | 93.8 | |
| SFTBackbone=Qwen2.5-7B-Instruct, Training=SFT, Few-shot=zero-shot2026.03 | 93.6 | |
| SFTBackbone=Qwen2.5-7B-Instruct, Training=SFT, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 93.6 | |
| Qwen2.5-3B-InstructBackbone=Qwen2.5-3B-Instruct, Few-shot=zero-shot2026.03 | 92 | |
| RLVRBackbone=Qwen2.5-3B-Instruct, Training=RLVR, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 91.8 | |
| MIPOBackbone=Qwen2.5-3B-Instruct, Training=MIPO, Few-shot=zero-shot2026.03 | 90.8 | |
| SFTBackbone=Qwen2.5-3B-Instruct, Training=SFT, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 90.33 | |
| SFTBackbone=Qwen2.5-3B-Instruct, Training=SFT, Few-shot=zero-shot2026.03 | 87.47 | |
| RLVRBackbone=Qwen2.5-1.5B-Instruct, Training=RLVR, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 85.4 | |
| MIPOBackbone=Llama-3.2-3B-Instruct, Training=MIPO, Few-shot=zero-shot2026.03 | 85.26 | |
| MIPOBackbone=Qwen2.5-1.5B-Instruct, Training=MIPO, Few-shot=zero-shot2026.03 | 82.27 | |
| Llama-3.2-3B-InstructBackbone=Llama-3.2-3B-Instruct, Few-shot=zero-shot2026.03 | 80.4 | |
| RLVRBackbone=Llama-3.2-3B-Instruct, Training=RLVR, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 80.07 | |
| Qwen2.5-1.5B-InstructBackbone=Qwen2.5-1.5B-Instruct, Few-shot=zero-shot2026.03 | 79 | |
| SFTBackbone=Llama-3.2-3B-Instruct, Training=SFT, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 78.73 | |
| SFTBackbone=Llama-3.2-3B-Instruct, Training=SFT, Few-shot=zero-shot2026.03 | 77.47 | |
| SFTBackbone=Qwen2.5-1.5B-Instruct, Training=SFT, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 75.2 | |
| SFTBackbone=Qwen2.5-1.5B-Instruct, Training=SFT, Few-shot=zero-shot2026.03 | 66.07 | |
| RLVRBackbone=Llama-3.2-1B-Instruct, Training=RLVR, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 58.8 | |
| MIPOBackbone=Llama-3.2-1B-Instruct, Training=MIPO, Few-shot=zero-shot2026.03 | 51.87 | |
| Llama-3.2-1B-InstructBackbone=Llama-3.2-1B-Instruct, Few-shot=zero-shot2026.03 | 44 | |
| SFTBackbone=Llama-3.2-1B-Instruct, Training=SFT, Few-shot=zero-shot2026.03 | 34.13 | |
| SFTBackbone=Llama-3.2-1B-Instruct, Training=SFT, Ground-truth usage=true, Few-shot=zero-shot2026.03 | 28.53 |