arXiv cs.AI· Miaobo Hu, Shuhao Hu, Xiaobo Guo, Xin Wang, Bokun Wang, Peng Zhang, Daren Zha·· 12 小时前AI 评分18
RSD-Poker：面向不完全信息博弈残差策略的结构自适应与偏移鲁棒风险-效用认证框架
RSD-Poker: Structure-Adaptive and Shift-Robust Risk-Utility Certification for Residual Policies in Imperfect-Information Games
AI 导读
RSD-Poker 提出一种结构自适应、偏移鲁棒的风险-效用认证框架，冻结残差族与尺度，在独立结构划分上学习策略可见分组并在校准前固定。在分层留出状态上，学习分区的对偶选择器将弱响应效用从全局对偶认证的 4.4074 提升至 4.4936，违规率从 0.0215 降至 0.0078，其混合鲁棒变体违规率为 0.0059。
正文 · 原文
Abstract:Residual policy adaptation provides a lightweight way to modify a strong reference policy, but a shared scale and a fixed subgroup partition can hide heterogeneous degradation and become fragile when the deployment mixture of information states changes. We introduce RSD-Poker, a structure-adaptive and shift-robust certification framework that freezes a bank of residual families and scales, learns a policy-visible partition on an independent structure split, and freezes that partition before calibration labels are joined. Each candidate-group pair receives a weighted simultaneous upper certificate for anchor-relative risk and a lower certificate for weak-response utility. A robust group-to-candidate map is then selected over a predeclared uncertainty set of deployment group proportions. Under independent calibration units drawn from each frozen group's law, a candidate bank and partition fixed before calibration, and invariant within-group conditionals, the selected map satisfies its declared mixture-robust risk budget and utility certificate with probability at least $1-\zeta_{risk}-\zeta_{util}$. The information contract supports both a teacher-backed transform and a teacher-free observation-only student. The retained deterministic 24-state audit remains an exact replay diagnostic: empirical-zero selects $\alpha=0.08$, raising the weak-response proxy from 4.2082 to 4.2889 with $0/12$ held-out threshold crossings. On stratified held-out states, the learned-partition dual selector raises weak utility from 4.4074 under global dual certification to 4.4936 and lowers held-out violation from 0.0215 to 0.0078; its mixture-robust variant reaches violation 0.0059. Across five observation-only checkpoints, risk-calibrated residuals attain weak utility $4.3659\pm0.0177$ and violation rate $0.0178\pm0.0057$.
| Comments: | 36 pages, 5 figures |
| Subjects: | Artificial Intelligence (cs.AI); Computer Science and Game Theory (cs.GT); Machine Learning (cs.LG) |
| Cite as: | arXiv:2609.33669 [cs.AI] |
| (or arXiv:2609.33669v2 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2609.33669 arXiv-issued DOI via DataCite |
Submission history
From: Miaobo Hu [view email]
[v1]
Sun, 27 Sep 2026 15:29:27 UTC (1,465 KB)
[v2]
Fri, 16 Oct 2026 11:41:51 UTC (1,465 KB)
来源：arXiv cs.AI · arxiv.org