Hide to See: Reasoning-prefix Masking for Visual-anchored Thinking in VLM Distillation

arXiv:2605.11651v2 Announce Type: replace-cross Abstract: Recent think-answer approaches in VLMs, such as Qwen3-VL-Thinking, boost reasoning performance by leveraging intermediate thinking steps before the final answer, but their high computational cost limits real-world deployment. To distill such...