BagelVLA
Original research
RESEARCH PAPER · Direct WAM
BagelVLA: Enhancing Long-Horizon Manipulation via Interleaved Vision-Language-Action Generation
Yucheng Hu, Jianke Zhang, Yuanfei Luo, et al.
Publication date2026-02-10
SourcearXiv
Research topicHybrid methods
Record reviewprimary record checked
Editorial context
A predictive vision-language-action architecture for multi-stage tasks.