Game of Thought: Robust Information Seeking with Large Language Models Using Game Theory

2026-02-02

AI Summary

This work addresses the limited robustness of large language models (LLMs) in actively acquiring information under conditions of insufficient data, where existing approaches suffer significant performance degradation in worst-case scenarios. The paper introduces the first formalization of adversarial information seeking as a strategic language search problem, proposing the Game of Thought framework. This framework models the information acquisition process as a two-player zero-sum extensive-form game and optimizes LLM behavior through strategies approximating Nash equilibrium. By integrating game theory, extensive-form game modeling, and prompt search, the method consistently and substantially improves worst-case performance on the Twenty Questions task, outperforming both direct prompting and heuristic search baselines.

Technology Category

• Search and Optimization: Adversarial Search
• Game Theory and Economic Paradigms: Adversarial Learning
• Multiagent Systems: Adversarial Agents

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Large Language Models (LLMs) are increasingly deployed in real-world scenarios where they may lack sufficient information to complete a given task. In such settings, the ability to actively seek out missing information becomes a critical capability. Existing approaches to enhancing this ability often rely on simplifying assumptions that degrade \textit{worst-case} performance. This is an issue with serious implications in high-stakes applications. In this work, we use the game of Twenty Questions to evaluate the information-seeking ability of LLMs. We introduce and formalize its adversarial counterpart, the Strategic Language Search (SLS) problem along with its variants as a two-player zero-sum extensive form game. We propose Game of Thought (GoT), a framework that applies game-theoretic techniques to approximate a Nash equilibrium (NE) strategy for the restricted variant of the game. Empirical results demonstrate that our approach consistently improves worst-case performance compared to (1) direct prompting-based methods and (2) heuristic-guided search methods across all tested settings.

Problem

• Large Language Models
• Information Seeking
• Worst-case Performance
• Adversarial Setting
• Strategic Language Search

Innovation

• Game of Thought
• Strategic Language Search
• Nash equilibrium
• adversarial information seeking
• large language models