Language Understanding on In-domain (test)
66.6AccuracyRLVR
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RLVRBackbone Model=Qwen2.5VL-7B-Instruct2026.07 | 66.6 | 0.514 | 0.285 | |
| RLCRBackbone Model=Qwen2.5VL-7B-Instruct2026.07 | 65.1 | 0.751 | 0.109 | |
| C3RL w/o RefBackbone Model=Qwen2.5VL-7B-Instruct2026.07 | 64.9 | 0.7 | 0.119 | |
| C3RLBackbone Model=Qwen2.5VL-7B-Instruct2026.07 | 64.4 | 0.748 | 0.076 | |
| SaySelfBackbone Model=Qwen2.5VL-7B-Instruct2026.07 | 55 | 0.807 | 0.075 | |
| SCBackbone Model=Qwen2.5VL-7B-Instruct2026.07 | 53.9 | 0.762 | 0.088 | |
| C3RLBackbone Model=Llama-3.2-3B-Instruct2026.07 | 49.1 | 0.734 | 0.142 | |
| SFT+RefBackbone Model=Qwen2.5VL-7B-Instruct2026.07 | 48 | 0.72 | 0.098 | |
| C3RL w/o RefBackbone Model=Llama-3.2-3B-Instruct2026.07 | 47.3 | 0.697 | 0.136 | |
| RLVRBackbone Model=Llama-3.2-3B-Instruct2026.07 | 45.9 | 0.587 | 0.463 | |
| BaseBackbone Model=Qwen2.5VL-7B-Instruct2026.07 | 45.3 | 0.531 | 0.448 | |
| RLCRBackbone Model=Llama-3.2-3B-Instruct2026.07 | 43.3 | 0.799 | 0.072 | |
| SCBackbone Model=Llama-3.2-3B-Instruct2026.07 | 40.8 | 0.787 | 0.091 | |
| SFT+RefBackbone Model=Llama-3.2-3B-Instruct2026.07 | 37.6 | 0.572 | 0.168 | |
| BaseBackbone Model=Llama-3.2-3B-Instruct2026.07 | 37.2 | 0.563 | 0.134 | |
| SaySelfBackbone Model=Llama-3.2-3B-Instruct2026.07 | 13.1 | 0.676 | 0.06 |