V-Zero: Answer-Label-Free On-Policy Distillation with Contrastive Evidence Gating for Fine-Grained Visual Reasoning
Paper • 2606.25319 • Published • 15
haoxiang sun PRO
hao05
AI & ML interests
None yet
Recent Activity
authored a paper about 1 hour ago
Dr. Seg: Revisiting GRPO Training for Visual Large Language Models through Perception-Oriented Design authored a paper about 2 hours ago
V-Zero: Answer-Label-Free On-Policy Distillation with Contrastive Evidence Gating for Fine-Grained Visual Reasoning updated a model about 6 hours ago
hao05/v-zero-4bOrganizations
None yet
upvoted a paper about 6 hours ago
upvoted an article 3 days ago
Article
V-Zero
hao05
• • 5 upvoted a paper 3 months ago