arXiv cs.CL· Hayden Arnold·· 13 小时前AI 评分31
面向 GLEE 的博弈论求解器：贝叶斯对手建模与道德修辞框架
Game-Theoretic Solvers with Bayesian Opponent Modeling and Moral Rhetoric Framing for GLEE
AI 导读
一个提交至 GLEE 竞赛的智能体，由经典博弈论求解器计算数值动作、在线贝叶斯层估计对手原型、语言模型撰写消息三部分组成，分别采用 Rubinstein 议价、Faratin 让步与 Kamenica–Gentzkow 披露模型，并用 Thompson sampling bandit 在五种道德基础框架上选择消息语域。
正文 · 原文
Abstract:We describe the agent we submitted to the GLEE Competition, which plays bargaining, negotiation, and persuasion games against a shared pool of human and AI opponents. The agent is built around three central components: a classical game-theoretic solver to compute numeric moves, an online Bayesian layer to estimate opponent archetypes, and a language model to write accompanying messages. We utilize Rubinstein bargaining for bargaining, Faratin concession for negotiation, and Kamenica--Gentzkow disclosure for persuasion, and adapt those models to live opponents using two separate layers: a Bayesian estimate of the opponent's hidden discount factor, reservation value, or signaling reliability, and a Thompson sampling bandit over five moral-foundation framings coupled with an evidence-intuition spectrum to establish message register. Most of the development effort went into the recurring problem that theoretically optimal policies perform poorly in certain live settings; we report associated failure modes, guardrails we built, and our hypotheses. We additionally evaluate our model over 13,200 Monte Carlo games against 11 synthetic opponent archetypes, an ablation of the opponent-modeling layer, and an extensive analysis of 9,337 live-game messages in which moral rhetoric framing in the persuasion setting is strongly correlated with whether the buyer purchased ($\chi^2=616.7$, $p \approx 3.7\times10^{-132}$), surviving controls for quality, round, and configuration.
| Comments: | 13 pages, 1 figure; accepted to 1st Workshop on Interpreting Agent Behavior (IAB) @ NeurIPS 2026 |
| Subjects: | Computer Science and Game Theory (cs.GT); Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.37397 [cs.GT] |
| (or arXiv:2610.37397v1 [cs.GT] for this version) | |
| https://doi.org/10.48550/arXiv.2610.37397 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Hayden Arnold [view email]
[v1]
Tue, 27 Oct 2026 15:17:04 UTC (54 KB)
来源：arXiv cs.CL · arxiv.org