Large language model (LLM) agents have shown remarkable progress in social deduction games (SDGs). However, existing approaches primarily focus on information processing and strategy selection, overlooking the significance of persuasive communication in influencing other players' beliefs and responses. In SDGs, success depends not only on making correct deductions but on convincing others to response in alignment with one's intent. To address this limitation, we formalize turn-based dialogue in SDGs as a Stackelberg competition, where the current player acts as the leader who strategically influences the follower's response. Building on this theoretical foundation, we propose a reinforcement learning framework that trains agents to optimize utterances for persuasive impact. Through comprehensive experiments across three diverse SDGs, we demonstrate that our agents significantly outperform baselines. This work represents a significant step toward developing AI agents capable of strategic social influence, with implications extending to scenarios requiring persuasive communication.

As a variant of the famous communication game Werewolf, One Night Ultimate W erewolf (ONUW) requires players to develop strategic discussion policies due to the potential role changes that increase the uncertainty and complexity of the game.

We investigate strategic deception in large language models using two complementary testbeds: Secret Agenda (across 38 models) and Insider Trading compliance (via SAE architectures). Secret Agenda reliably induced lying when deception advantaged goal achievement across all model families. Analysis revealed that autolabeled SAE features for "deception" rarely activated during strategic dishonesty, and feature steering experiments across 100+ deception-related features failed to prevent lying. Conversely, insider trading analysis using unlabeled SAE activations separated deceptive versus compliant responses through discriminative patterns in heatmaps and t-SNE visualizations. These findings suggest autolabel-driven interpretability approaches fail to detect or control behavioral deception, while aggregate unlabeled activations provide population-level structure for risk assessment. Results span Llama 8B/70B SAE implementations and GemmaScope under resource constraints, representing preliminary findings that motivate larger studies on feature discovery, labeling methodology, and causal interventions in realistic deception contexts.

Achieving Artificial General Intelligence (AGI) requires AI agents that can not only make stratigic decisions but also engage in flexible and meaningful communication. Inspired by Wittgenstein's language game theory in Philosophical Investigations, we propose that language agents can learn through in-context interaction rather than traditional multi-stage frameworks that separate decision-making from language expression. Using Werewolf, a social deduction game that tests language understanding, strategic interaction, and adaptability, we develop the Multi-agent Kahneman & Tversky's Optimization (MaKTO). MaKTO engages diverse models in extensive gameplay to generate unpaired desirable and unacceptable responses, then employs KTO to refine the model's decision-making process. In 9-player Werewolf games, MaKTO achieves a 61% average win rate across various models, outperforming GPT-4o and two-stage RL agents by relative improvements of 23.0% and 10.9%, respectively. Notably, MaKTO also demonstrates human-like performance, winning 60% against expert players and showing only 49% detectability in Turing-style blind tests. These results showcase MaKTO's superior decision-making, strategic adaptation, and natural language generation in complex social deduction games.

Trustworthiness reasoning is crucial in multiplayer games with incomplete information, enabling agents to identify potential allies and adversaries, thereby enhancing reasoning and decision-making processes. Traditional approaches relying on pre-trained models necessitate extensive domain-specific data and considerable reward feedback, with their lack of real-time adaptability hindering their effectiveness in dynamic environments. In this paper, we introduce the Graph Retrieval Augmented Reasoning (GRATR) framework, leveraging the Retrieval-Augmented Generation (RAG) technique to bolster trustworthiness reasoning in agents. GRATR constructs a dynamic trustworthiness graph, updating it in real-time with evidential information, and retrieves relevant trust data to augment the reasoning capabilities of Large Language Models (LLMs). We validate our approach through experiments on the multiplayer game "Werewolf," comparing GRATR against baseline LLM and LLM enhanced with Native RAG and Rerank RAG. Our results demonstrate that GRATR surpasses the baseline methods by over 30\% in winning rate, with superior reasoning performance. Moreover, GRATR effectively mitigates LLM hallucinations, such as identity and objective amnesia, and crucially, it renders the reasoning process more transparent and traceable through the use of the trustworthiness graph.

Strategic social deduction games serve as valuable testbeds for evaluating the understanding and inference skills of language models, offering crucial insights into social science, artificial intelligence, and strategic gaming. This paper focuses on creating proxies of human behavior in simulated environments, with Among Us utilized as a tool for studying simulated human behavior. The study introduces a text-based game environment, named AmongAgents, that mirrors the dynamics of Among Us. Players act as crew members aboard a spaceship, tasked with identifying impostors who are sabotaging the ship and eliminating the crew. Within this environment, the behavior of simulated language agents is analyzed. The experiments involve diverse game sequences featuring different configurations of Crewmates and Impostor personality archetypes. Our work demonstrates that state-of-the-art large language models (LLMs) can effectively grasp the game rules and make decisions based on the current context. This work aims to promote further exploration of LLMs in goal-oriented games with incomplete information and complex action spaces, as these settings offer valuable opportunities to assess language model performance in socially driven scenarios.

Large language models (LLMs) have exhibited memorable strategic behaviors in social deductive games. However, the significance of opinion leadership exhibited by LLM-based agents has been overlooked, which is crucial for practical applications in multi-agent and human-AI interaction settings. Opinion leaders are individuals who have a noticeable impact on the beliefs and behaviors of others within a social group. In this work, we employ the Werewolf game as a simulation platform to assess the opinion leadership of LLMs. The game features the role of the Sheriff, tasked with summarizing arguments and recommending decision options, and therefore serves as a credible proxy for an opinion leader. We develop a framework integrating the Sheriff role and devise two novel metrics for evaluation based on the critical characteristics of opinion leaders. The first metric measures the reliability of the opinion leader, and the second assesses the influence of the opinion leader on other players' decisions. We conduct extensive experiments to evaluate LLMs of different scales. In addition, we collect a Werewolf question-answering dataset (WWQA) to assess and enhance LLM's grasp of the game rules, and we also incorporate human participants for further analysis. The results suggest that the Werewolf game is a suitable test bed to evaluate the opinion leadership of LLMs and few LLMs possess the capacity for opinion leadership.

This paper presents an innovative framework that integrates Large Language Models (LLMs) with an external Thinker module to enhance the reasoning capabilities of LLM-based agents. Unlike augmenting LLMs with prompt engineering, Thinker directly harnesses knowledge from databases and employs various optimization techniques. The framework forms a reasoning hierarchy where LLMs handle intuitive System-1 tasks such as natural language processing, while the Thinker focuses on cognitive System-2 tasks that require complex logical analysis and domain-specific knowledge. Our framework is presented using a 9-player Werewolf game that demands dual-system reasoning. We introduce a communication protocol between LLMs and the Thinker, and train the Thinker using data from 18800 human sessions and reinforcement learning. Experiments demonstrate the framework's effectiveness in deductive reasoning, speech generation, and online game evaluation. Additionally, we fine-tune a 6B LLM to surpass GPT4 when integrated with the Thinker. This paper also contributes the largest dataset for social deduction games to date.

In this paper, we explore the potential of Large Language Models (LLMs) Agents in playing the strategic social deduction game, Resistance Avalon. Players in Avalon are challenged not only to make informed decisions based on dynamically evolving game phases, but also to engage in discussions where they must deceive, deduce, and negotiate with other players. These characteristics make Avalon a compelling test-bed to study the decision-making and language-processing capabilities of LLM Agents. To facilitate research in this line, we introduce AvalonBench - a comprehensive game environment tailored for evaluating multi-agent LLM Agents. This benchmark incorporates: (1) a game environment for Avalon, (2) rule-based bots as baseline opponents, and (3) ReAct-style LLM agents with tailored prompts for each role. Notably, our evaluations based on AvalonBench highlight a clear capability gap. For instance, models like ChatGPT playing good-role got a win rate of 22.2% against rule-based bots playing evil, while good-role bot achieves 38.2% win rate in the same setting. We envision AvalonBench could be a good test-bed for developing more advanced LLMs (with self-playing) and agent frameworks that can effectively model the layered complexities of such game environments.