arXiv preprint / cs.CV

MMVistaReason: Toward Open-Data and Post-Training Recipes for Multimodal Reasoning

作者：Juekai Lin, Honglin Lin, Yuqian Yuan, Xiaolong Wu, Jie Cao, Liang Liang, Yunqi Cao, Yun Zhu, Wenqiao Zhang, Lijun Wu · 会议/期刊：arXiv preprint · 方向：cs.CV · 发布日期：2026-10-02

2026-10-02视觉语言
新标签打开 AlphaXiv
MMVistaReason: Toward Open-Data and Post-Training Recipes for Multimodal Reasoning figure
AlphaXiv 中文概览（可滚动查看）