@misc{indiciae292c460d5c74, title = {Grounding the Score: Explicit Visual Premise Verification for Reliable Vision-Language Process Reward Models}, author = {Junxin Wang and Dai Guan and Weijie Qiu and Zhihang Li and Yongbo Gai and Zhengyi Yang and Mengyu Zhou and Erchao Zhao and Xiaoxi Jiang and Guanjun Jiang}, year = {2026}, url = {https://arxiv.org/abs/2603.16253}, note = {Source identifier: 2603.16253} }