New arXiv papers explore multi-agent AI systems

Eight recent arXiv papers investigate multi-agent AI, covering norm enforcement, feedback control, and agentic reasoning. Research explores how AI agents interact, form conventions, and can be verified in production environments.