Visual Perception on Hallusion
72.45AccuracyQwen2.5-VL-32B-Instruct + NoisyRollout
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen2.5-VL-32B-Instruct + NoisyRolloutStudent Model=Qwen2.5-VL-32B-Instruct, Distillation/Training Method=NoisyRollout2026.03 | 72.45 | |
| NoisyRolloutTeacher Model=Qwen2.5-VL-32B-Instruct + NoisyRollout, Student Model=Qwen2.5-VL-7B-Instruct, Distillation/Training Method=NoisyRollout2026.03 | 70.66 | |
| REOPOLDTeacher Model=Qwen2.5-VL-32B-Instruct + NoisyRollout, Student Model=Qwen2.5-VL-7B-Instruct, Distillation/Training Method=REOPOLD2026.03 | 70.14 | |
| GRPOTeacher Model=Qwen2.5-VL-32B-Instruct + NoisyRollout, Student Model=Qwen2.5-VL-7B-Instruct, Distillation/Training Method=GRPO2026.03 | 69.8 | |
| RKLTeacher Model=Qwen2.5-VL-32B-Instruct + NoisyRollout, Student Model=Qwen2.5-VL-7B-Instruct, Distillation/Training Method=RKL2026.03 | 69.51 | |
| Qwen2.5-VL-7B-InstructStudent Model=Qwen2.5-VL-7B-Instruct2026.03 | 65.62 | |
| REOPOLDTeacher Model=Qwen2.5-VL-32B-Instruct + NoisyRollout, Student Model=Qwen2.5-VL-3B-Instruct, Distillation/Training Method=REOPOLD2026.03 | 63.62 | |
| PAPOTeacher Model=Qwen2.5-VL-32B-Instruct + NoisyRollout, Student Model=Qwen2.5-VL-3B-Instruct, Distillation/Training Method=PAPO2026.03 | 61.62 | |
| Qwen2.5-VL-3B-InstructStudent Model=Qwen2.5-VL-3B-Instruct2026.03 | 61.51 | |
| RKLTeacher Model=Qwen2.5-VL-32B-Instruct + NoisyRollout, Student Model=Qwen2.5-VL-3B-Instruct, Distillation/Training Method=RKL2026.03 | 60.83 |