Call for Participation Shared Task for the 2nd Workshop of AI Werewolf and Dialog System (AIWolfDial2024) at the 17th International Natural Language Generation conference (INLG 2024) # Summary Recent achievements of generation models, e.g. ChatGPT, are gathering greater attentions. However, there is still room to investigate LLMs could sufficiently able to handle coherent responses, longer contexts, common grounds, and logics. Werewolf is a social, hidden identity game that requires debate between players and coalition building. The goal of our AIWerewolf contest is to build an AI agent that is able to play this game against other AI. # Schedule Shared tasks July 28th, 2024 Registration August 4th, 2024 Preliminary run (self-match game) mid August, 2024 Formal run (multi-agent game) Workshop August 18th, 2024 Paper submission deadline (submissions should be via the Sontconf system, see our Call for Papers) August 25th, 2024 Notifications of the paper accpetance August 30th, 2024 Camera ready paper deadline Sep 24th, 2024 Workshop (planned in pm) in Tokyo Sep 23-27, 2024 INLG conference Our shared task is held as a part of our AIWolfDial 2024 workshop at INLG 2024 (17th International Natural Language Generation Conference), which will be held in Tokyo from September 23th to 27th. It is not mandatry for our shared task participants to attend the INLG 2024 conference, but encouraged to submit thier papers to the workshop. Please refer to our websites for the details including technical requirments: https://sites.google.com/view/aiwolfdial2024-inlg We have a seperate call for papers of our workshop. # Why AI Werewolf? Recent achievements of generation models, e.g. ChatGPT, are gathering greater attentions. However, such a huge language model would not be sufficiently able to handle coherent responses, longer contexts, common grounds, and logics. The AIWolfDial 2024 contest, which is an international open contest for automatic players of the conversation game "Mafia", requires players not just to communicate but to infer, persuade, deceive other players via coherent logical conversations, while having the role-playing non-task-oriented chats as well. We believe that this contest reveals current issues in the recent huge language models, showing directions of next breakthrough in the NLP area. From the viewpoint of Game AI area, players must hide information, in contrast to perfect information games such as chess or Reversi. Each player acquires secret information from other players' conversations and behavior and acts by hiding information to accomplish their objectives. Players are required persuasion for earning confidence, and speculation for detecting fabrications. Participants must build an artificial intelligence agent that can play the werewolf game as humans do, using natural language. Participant agents will be evaluated by a panel of judges, who will grade the subjective quality of the dialog generated by the agent, in addition to their win rates. Agents must communicate in Japanese or English. # Registration A team should send a mail to aiwolf [at] kanolab.net (replace at by @), describing your team name, a contact e-mail address, names and affiliations of its members (please mark a contact person when a team consists of multiple members), communication language (English and/or Japanese) of your agent, ssh public key and your preferred user name to connect to our game server. Registration is free. # System Evaluation Participants should submit a paper to the workshop, or a system design description document to the organizers. In addition to the win rates, reviewers will perform subjective evaluations on the game logs of a self-match games and multi-agent games, using following criteria: A Natural utterance expressions B Contextually natural conversation C Coherent (not contradictory) conversation D Coherent game actions (vote, attack, divine) with conversation contents E Diverse utterance expressions, including coherent characterization Please note that vague utterances that could be used regardless of context are not always natural in the werewolf game. The top-ranking teams will be awarded prizes and gifts from SpiralAI, a company developing its own LLM for colloquial multi-turn conversations. # Sponser Spiral.AI Inc, Japan # Organizers Organizers and Program Commitee: Yoshinobu Kano, Shizuoka University, Japan Claus Aranha, Tsukuba University Takashi Otsuki, Yamagata University, Japan Fujio Toriumi, The University of Tokyo, Japan Hirotaka Osawa, Keio University, Japan Daisuke Katagami, Tokyo Polytechnic University, Japan Michimasa Inaba, The University of Electro-Communications, Japan Kei Harada, The University of Electro-Communications, Japan Takeshi Ito, The University of Electro-Communications, Japan Local Organizers: Neo Watanabe, Shizuoka University, Japan Kaito Kagaminuma, Shizuoka University, Japan Yuto Sahashi, Shizuoka University, Japan On behalf of the AIWolf organizers, Yoshinobu Kano Associate Professor, Shizuoka University kano@inf.shizuoka.ac.jp