Summary of Learning to Discuss Strategically: a Case Study on One Night Ultimate Werewolf, by Xuanfa Jin et al.
Learning to Discuss Strategically: A Case Study on One Night Ultimate Werewolf
by Xuanfa Jin, Ziyan Wang, Yali Du, Meng Fang, Haifeng Zhang, Jun Wang
First submitted to arxiv on: 30 May 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel approach to strategic communication in large language models is proposed, addressing the limitations of recent agents that neglect control over discussion tactics. The research focuses on One Night Ultimate Werewolf, a variant of the famous communication game Werewolf, where players must develop strategic discussion policies due to role changes and uncertainty. In two scenarios, Perfect Bayesian Equilibria (PBEs) are shown to exist, highlighting the significance of discussion tactics in affecting players’ utilities. A reinforcement learning-instructed language agent framework is proposed, using a trained discussion policy to determine appropriate discussion tactics. Experimental results demonstrate the effectiveness and generalizability of the framework. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Imagine you’re playing a game where you need to figure out who’s trying to trick you! That’s what One Night Ultimate Werewolf is all about. Researchers studied this game to understand how people communicate and make decisions. They found that when people talk, it really matters what they say and how they say it. This can actually change the outcome of the game! To help players make better choices, the researchers created a special computer program that learns from experience. It can pick up on clues and adjust its strategy to win the game. The results show that this program is pretty good at playing the game, and it might even be useful in real-life situations where people need to work together or negotiate. |
Keywords
» Artificial intelligence » Reinforcement learning
