arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2510.18567cs.GT
收藏

委托-代理博弈中的上下文搜索:退化的诅咒

Contextual Search in Principal-Agent Games: The Curse of Degeneracy

Hong Kong University of Science and Technology(香港科技大学);Central China Normal University(华中师范大学);Shanghai University of Finance and Economics(上海财经大学);Great Bay University(大湾区大学)

Yiding Feng, Mengfan Ma, Bo Peng, Zongqi Wan

AI总结:

本文研究一般委托-代理博弈中的上下文搜索,证明悲观与经典 Stackelberg 遗憾的指数及下界,并将难度激增归因于上下文行动退化。

摘要:

在本工作中,我们引入并研究一般委托-代理博弈中的上下文搜索,其中委托人通过基于上下文信息和历史反馈提供合同来反复与代理人互动,而不知道代理人的真实成本或奖励。我们的模型通过容纳更丰富的代理人行动空间,推广了经典的上下文定价。在具有 d 维上下文的 T 轮中,我们针对悲观 Stackelberg 遗憾建立了一个渐近紧的指数 T^{1−Θ(1/d)} 界,其基准是与观察到的反馈相一致的委托人最佳效用。我们还建立了委托-代理博弈经典 Stackelberg 遗憾的 Ω(T^{1/2−1/(2d)}) 下界,展示了与上下文定价问题(即具有两个行动的委托-代理博弈)之间令人惊讶的双指数难度分离;已知上下文定价问题具有近乎最优的 O(d log log T) 遗憾界 [Kleinberg and Leighton, 2003, Leme and Schneider, 2018, Liu et al., 2021]。特别地,这种双指数难度分离甚至发生在具有三个行动和二维上下文的特殊情形中。我们发现,这种学习难度的显著增加源于一种我们称为上下文行动退化的结构现象:对抗性选择的上下文可能使某些行动被严格占优(因此无法被激励),阻碍委托人探索或了解这些行动的能力,并从根本上限制学习进展。

英文摘要:

In this work, we introduce and study contextual search in general principal-agent games, where a principal repeatedly interacts with agents by offering contracts based on contextual information and historical feedback, without knowing the agents' true costs or rewards. Our model generalizes classical contextual pricing by accommodating richer agent action spaces. Over $T$ rounds with $d$-dimensional contexts, we establish an asymptotically tight exponential $T^{1 - Θ(1/d)}$ bound in terms of the pessimistic Stackelberg regret, benchmarked against the best utility for the principal that is consistent with the observed feedback. We also establish a lower bound of $Ω(T^{\frac{1}{2}-\frac{1}{2d}})$ on the classic Stackelberg regret for principal-agent games, demonstrating a surprising double-exponential hardness separation from the contextual pricing problem (a.k.a, the principal-agent game with two actions), which is known to admit a near-optimal $O(d\log\log T)$ regret bound [Kleinberg and Leighton, 2003, Leme and Schneider, 2018, Liu et al., 2021]. In particular, this double-exponential hardness separation occurs even in the special case with three actions and two-dimensional context. We identify that this significant increase in learning difficulty arises from a structural phenomenon that we call contextual action degeneracy, where adversarially chosen contexts can make some actions strictly dominated (and hence unincentivizable), blocking the principal's ability to explore or learn about them, and fundamentally limiting learning progress.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑