Exploring Large Language Models for Communication Games: An Empirical Study on Werewolf · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2309.04658/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTExploring Large Language Models for Communication Games: An Empirical Study on Werewolf — mentioned by sabrina_ramonovhttps://news.ycombinator.com/item?id=41045704urn:hackerpapers:mention:2309.04658:41045704Tue, 23 Jul 2024 13:18:51 GMT<p>&quot;This repository is the code for the paper &quot;Exploring Large Language Models for Communication Games: An Empirical Study on Werewolf&quot;. In this paper, we explore the problem of how to engage large language models (LLMs) in communication games, and in response, propose a tuning-free framework. Our approach keeps LLMs frozen, and relies on the retrieval and reflection on past communications and experiences for improvement. An empirical study on the representative and widely-studied communication game, &quot;Werewolf&quot;, demonstrates that our framework can effectively play Werewolf game without tuning…</p><p><a href="https://hackerpaper.kaipfister.de/p/2309.04658/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.CL