Josef Dai

Researcher, Peking University

2 papers at NeurIPS 2025

Homepage· OpenReview· Semantic Scholar· Google Scholar

Poster Session 4

1 paper

Thursday, December 4, 2025 · 4:30 PM → 7:30 PM

Exhibit Hall C,D,E

SafeVLA: Towards Safety Alignment of Vision-Language-Action Model via Constrained Learning

#1103 Spotlight · Borong Zhang, Yuhao Zhang, Jiaming Ji, Yingshan Lei, Josef Dai, Yuanpei Chen, Yaodong Yang

We make Vision-Language-Action models (VLAs) significantly safer (83.58% improvement, no performance loss) by explicitly integrating constraints via a new integrated safety approach (ISA) based on CMDPs/SafeRL, validated on our new benchmark.

Poster Session 5

1 paper

Friday, December 5, 2025 · 11:00 AM → 2:00 PM

Exhibit Hall C,D,E

InterMT: Multi-Turn Interleaved Preference Alignment with Human Feedback

#5104 Spotlight · Boyuan Chen, Donghai Hong, Jiaming Ji, Jiacheng Zheng, Bowen Dong, Jiayi Zhou, Kaile Wang, Josef Dai, Xuyao Wang, Wenqi Chen, Qirui Zheng, Wenxin Li, Sirui Han, Yike Guo, Yaodong Yang

A human preference dataset for multi-turn interleaved multimodal understanding and generatin tasks