mlsanityData to 7 Oct 2026

DVM: Towards Controllable LLM Agents in Social Deduction Games

arXiv 2501.06695 ↗First posted 12 Jan 2025cs.AIcited by 6

On the map: LLM agents and tool use → Reinforcement learning for LLM agents

Abstract (arXiv)

Large Language Models (LLMs) have advanced the capability of game agents in social deduction games (SDGs). These games rely heavily on conversation-driven interactions and require agents to infer, make decisions, and express based on such information. While this progress leads to more sophisticated and strategic non-player characters (NPCs) in SDGs, there exists a need to control the proficiency of these agents. This control not only ensures that NPCs can adapt to varying difficulty levels during gameplay, but also provides insights into the safety and fairness of LLM agents. In this paper, we present DVM, a novel framework for developing controllable LLM agents for SDGs, and demonstrate its implementation on one of the most popular SDGs, Werewolf. DVM comprises three main components: Predictor, Decider, and Discussor. By integrating reinforcement learning with a win rate-constrained decision chain reward mechanism, we enable agents to dynamically adjust their gameplay proficiency to achieve specified win rates. Experiments show that DVM not only outperforms existing methods in the Werewolf game, but also successfully modulates its performance levels to meet predefined win rate targets. These results pave the way for LLM agents' adaptive and balanced gameplay in SDGs, opening new avenues for research in controllable game agents.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Placed in Reinforcement learning for LLM agents, one of 16 directions in LLM agents and tool use. Builds on 25 of its 32 references that are in the map. First posted on arXiv 12 Jan 2025.

Builds on

25 of 32 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024A Survey on Large Language Model-Based Game Agents2 Apr
  2. 2024Enhance Reasoning for Large Language Models in the Game Werewolf4 Feb
  3. 2023Language Agents with Reinforcement Learning for Strategic Play in the Werewolf Game29 Oct
  4. 2023Voyager: An Open-Ended Embodied Agent with Large Language Models25 May

Other references · 21

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2017Proximal Policy Optimization Algorithmscited by 14,441
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  4. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  5. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  6. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,304
  7. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Modelscited by 1,397
  8. 2024ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Toolscited by 1,096
  9. 2024Is DPO Superior to PPO for LLM Alignment? A Comprehensive Studycited by 262
  10. 2023Exploring Large Language Models for Communication Games: An Empirical Study on Werewolfcited by 238
  11. 2023Secrets of RLHF in Large Language Models Part I: PPOcited by 226
  12. 2023Playing repeated games with Large Language Modelscited by 213
9 more
  1. 2024Cradle: Empowering Foundation Agents Towards General Computer Controlcited by 114
  2. 2023MP5: A Multi-modal Open-ended Embodied System in Minecraft via Active Perceptioncited by 67
  3. 2024Determinants of LLM-assisted Decision-Makingcited by 55
  4. 2019Finding Friend and Foe in Multi-Agent Gamescited by 42
  5. 2023A Theoretical Analysis of Optimistic Proximal Policy Optimization in Linear Markov Decision Processescited by 28
  6. 2024AMONGAGENTS: Evaluating Large Language Models in the Interactive Text-Based Social Deduction Gamecited by 26
  7. 2024Baba Is AI: Break the Rules to Beat the Benchmarkcited by 10
  8. 2020Proximal Policy Optimization via Enhanced Exploration Efficiencycited by 9
  9. 2019Proximal Policy Optimization with Mixed Distributed Trainingcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 4; 2026: 2.

6 papers.

  1. 2025A Comprehensive Survey on Multi-Agent Cooperative Decision-Making: Scenarios, Approaches, Challenges and Perspectivescited by 27
  2. 2025VistaWise: Building Cost-Effective Agent with Cross-Modal Knowledge Graph for Minecraftcited by 13
  3. 2025The Stackelberg Speaker: Optimizing Persuasive Communication in Social Deduction Gamescited by 3
  4. 2025TongSIM: A General Platform for Simulating Intelligent Machinescited by 3
  5. 2026CaM-Wolf: Causal-Aware Multimodal Agents for Social Deduction Gamescited by 0
  6. 2026Adversarial Closed-Loop Curriculum for Evolving Role-Playing Agentscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Large Language Models as Agents in Two-Player Games
  2. 2024A Survey on Large Language Model-Based Social Agents in Game-Theoretic Scenarios
  3. 2024BALROG: Benchmarking Agentic LLM and VLM Reasoning On Games
  4. 2024STRIDE: A Tool-Assisted LLM Agent Framework for Strategic and Interactive Decision-Making
  5. 2026Playing social deduction games with reinforcement fine-tuned large language models
  6. 2025Understanding LLM Agent Behaviours via Game Theory: Strategy Recognition, Biases and Multi-Agent Dynamics
  7. 2026Social Gym and SPaRTan: Benchmarking and Improving LLM Social Reasoning via Multi-Agent Game Tournaments
  8. 2023AvalonBench: Evaluating LLMs Playing the Game of Avalon