MACL deploys image, text, name, and coordination agents with message passing and adaptive balancing to achieve 1-5% precision gains on VISTA-Beyond for few-shot and zero-shot OOD alignment.
Human cognition involves the dynamic integra- tion of neural activity and neuromodulatory systems
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.MA 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Multi-Agent Cooperative Learning for Robust Vision-Language Alignment under OOD Concepts
MACL deploys image, text, name, and coordination agents with message passing and adaptive balancing to achieve 1-5% precision gains on VISTA-Beyond for few-shot and zero-shot OOD alignment.