1

Deep Pre-Alignment for VLMs
通过模块化视觉编码器替换实现更深层的多模态预对齐。
Tianyu Yu, Kechen Fang, Zihao Wan, Kaidong Zhang, Yicheng Zhang, Jun Song, Bo Zheng, Yuan Yao ICML 2026
Paper Code
MiniCPM-V 4.5: Cooking Efficient MLLMs via Architecture, Data, and Training Recipe
覆盖架构、数据与训练策略的高效多模态大模型构建方案。
Tianyu Yu, Zefan Wang, Chongyi Wang, Fuwei Huang, Wenshuo Ma, Zhihui He, +26 more authors, Zhiyuan Liu, Maosong Sun CVPR 2026
Paper Code
RLAIF-V: Open-Source AI Feedback Leads to Super GPT-4V Trustworthiness
用于可信多模态对齐的开源反馈数据。
Tianyu Yu, Haoye Zhang, Qiming Li, Qixin Xu, Yuan Yao, Da Chen, +8 more authors, Tat-Seng Chua, Maosong Sun CVPR 2025 Highlight
Paper Code
RLHF-V: Towards Trustworthy MLLMs via Behavior Alignment from Fine-grained Correctional Human Feedback
用细粒度纠错式人类反馈减少多模态大模型幻觉。
Tianyu Yu, Yuan Yao, Haoye Zhang, Taiwen He, Yifeng Han, Ganqu Cui, +3 more authors, Maosong Sun, Tat-Seng Chua CVPR 2024
Paper Code
Cross-Modal Omni Interaction Modeling for Phrase Grounding
用于短语定位的跨模态交互建模。
Tianyu Yu, Tianrui Hui, Zhihao Yu, Yue Liao, Sansi Yu, Faxi Zhang, Si Liu ACM MM 2020
Paper