Yaozhi ZHENG

VR-Thinker: Boosting Multimodal Reward Models through Think with Image Reasoning

Recent advancements in multimodal reward models (RMs) have substantially improved post-training for visual generative models. However, current RMs face inherent limitations: **(1)** visual inputs consume large context budgets, forcing fewer frames and causing a loss of details; and **(2)** all visua…