@article{differencefeedbackgeneratingmultimodalpr, title = {Difference Feedback: Generating Multimodal Process-Level Supervision for VLM Reinforcement Learning}, author = { Feiding and Yongkang Zhang and Yuhao Liao and Zijian Zeng and Chunzheng Zhu and Yaozong Zheng and Yafei Liu and Yeling Peng and Youwei Wang and Sibo Wang and Huiming Yang and Linglin Liao and Shunzhi Yang}, year = {2026}, eprint = {2603.27482}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2603.27482}, }