1
V-Zero: Answer-Label-Free On-Policy Distillation with Contrastive Evidence Gating for Fine-Grained Visual Reasoning
不靠标准答案也能蒸馏？对比证据门控加持，让细粒度视觉推理在无标签场景下自举优化，高效又亮眼。
arXiv:2606.25319v1 Announce Type: new Abstract: Fine-grained visual reasoning requires multimodal large language models (MLLMs) to identify task-relev…