Adaptive Collaboration with Humans: Metacognitive Policy Optimization for Multi-Agent LLMs with Continual Learning

Adaptive Collaboration with Humans: Metacognitive Policy Optimization for Multi-Agent LLMs with Continual Learning explores HILA enables multi-agent LLMs to collaborate with humans by learning when to solve problems autonomously and when to defer to human experts, improving performance on challenging tasks.. Commercial viability score: 7/10 in Multi-Agent LLMs.