TY  - RPRT
TI  - Multi-Agent Cooperative Learning for Robust Vision-Language Alignment under OOD Concepts
AU  - Philip Xu
PY  - 2026
UR  - https://arxiv.org/abs/2601.09746
ID  - 2601.09746
ER  -