Advancements in Multi-Agent Systems and Theory of Mind

The field of artificial intelligence is witnessing significant developments in multi-agent systems and theory of mind. Researchers are exploring new approaches to improve the collaboration and cooperation among agents, enabling them to better understand and reason about the intentions and mental states of others. This is crucial for effective human-AI interaction and has broad implications for applications such as translation, summarization, and cybersecurity. The integration of cognitive-theoretic insights and multi-agent reinforcement learning is leading to more advanced and adaptable systems. Notable papers in this area include TACTIC, which proposes a cognitively informed multi-agent framework for translation, and Agentic Neural Networks, which introduces a self-evolving multi-agent system via textual backpropagation. Additionally, UniToMBench provides a unified benchmark for improving and assessing theory of mind capabilities in large language models.

Sources

A MARL-based Approach for Easing MAS Organization Engineering

Let's Put Ourselves in Sally's Shoes: Shoes-of-Others Prefixing Improves Theory of Mind in Large Language Models

TACTIC: Translation Agents with Cognitive-Theoretic Interactive Collaboration

Agentic Neural Networks: Self-Evolving Multi-Agent Systems via Textual Backpropagation

Multi-Agent Language Models: Advancing Cooperation, Coordination, and Adaptation

UniToMBench: Integrating Perspective-Taking to Improve Theory of Mind in LLMs

Specification and Evaluation of Multi-Agent LLM Systems -- Prototype and Cybersecurity Applications

Built with on top of