Negotiation Game

Negotiation games, abstract models of bargaining interactions, are a burgeoning area of research focusing on how artificial agents can effectively negotiate and cooperate, or compete, with each other and humans. Current work explores various algorithms, including Bayesian learning, reinforcement learning with opponent shaping (like Advantage Alignment), and game-theoretic approaches such as Nash bargaining solutions, to optimize agent strategies in diverse negotiation scenarios. These studies leverage large language models (LLMs) as agents, evaluating their performance and alignment with human behavior in both cooperative and competitive settings, and investigating the impact of factors like emotions and information asymmetry. This research is significant for advancing AI safety and improving the design of AI systems that can effectively interact with humans in complex, real-world situations.