今日速览
| 序号 | 标题 | 来源 | 日期 | 主题 | 推荐等级 |
|---|---|---|---|---|---|
| 1 | AV-AIVAT: 74x Cheaper Agent Evaluation with Certified Anytime-Valid Stopping in Imperfect-Information Games | arXiv | 2026-08-06 | AI-Agent | 高 |
| 2 | Resourced Authority A Mechanism-Design Model for Participatory Governance of Deployed AI Agents | arXiv | 2026-08-06 | AI-Agent | 高 |
重点论文与技术动态
1. AV-AIVAT: 74x Cheaper Agent Evaluation with Certified Anytime-Valid Stopping in Imperfect-Information Games
- 来源：arXiv
- 日期：2026-08-06
- 作者/机构：Boning Li, Yu Chen, Longbo Huang
- 主题标签：
AI-Agent,arXiv - 推荐等级：高
- 分类：cs.GT, cs.AI, cs.CL, cs.LG, cs.MA
一句话结论
AV-AIVAT通过整合AIVAT方差减少与连续置信序列，实现智能体评估的高效、可审计早期停止，显著降低评估成本。
核心内容
- AIVAT利用条件均值零修正在15个LLM智能体配置中降低不完美信息博弈方差，中位数54倍，基于71,439对HUNL手牌测试。
- AV-AIVAT结合AIVAT与连续监控置信序列（CSs），确保停止时证据充分且置信水平有效，避免固定预算导致的资源浪费。
- 在95%置信水平和±1大盲注精度下，AV-AIVAT比原始方法节省74倍游戏数量（AsympCS），EB-CS实现1.37倍停止时间优势。
方法与数据
- 方法：采用AIVAT校正方差，并应用AsympCS和EB-CS进行实时停止决策，确保在线值模型仅依赖历史数据。
- 数据：实验覆盖71,439对HUNL手牌，15个LLM智能体配置，验证了方差减少与停止效率。
价值判断
- 值得关注：提供可验证的评估框架，允许评估在证据充分时立即停止，避免不必要开销。
- 可复用点：AIVAT的方差减少技术可迁移至其他博弈场景；CSs框架支持在线评估系统设计。
- 局限/待核查：EB-CS需独立验证校正收益边界，可能在非结构化博弈中增加实现复杂度。
摘要
Deciding which of two agents is stronger means playing games until skill outweighs luck, and every game costs money, model inference, or expert time. Since the number of games needed is unknown, fixed-budget evaluations either keep paying after the result is settled or stop before the agents can be told apart, while naive optional stopping with an ordinary confidence interval invalidates the stated level. We make such an evaluation stop as soon as its evidence suffices, with the guarantee intact. The Action-Informed Value Assessment Tool (AIVAT) reduces variance in imperfect-information games through conditional mean-zero corrections, by a median $54\times$ across 15 LLM agent configurations spanning 71,439 paired Heads-Up No-Limit Hold'em (HUNL) hands, but does not say when to stop. We combine AIVAT with continuously monitored Confidence Sequences (CSs) into anytime-valid AIVAT (AV-AIVAT), whose online value model learns only from past games so that no game scores its own correction. At the nominal 95\% level and a target precision of $\pm1$ Big Blind, raw outcomes need a median $74\times$ as many hands as AIVAT-corrected outcomes to stop under the Asymptotic CS (AsympCS). Exact finite-sample certification uses the Empirical-Bernstein CS (EB-CS), which needs an independently justified bound on corrected payoffs. We establish such a bound structurally for Leduc hold'em and characterize a width floor set by the CS's bet cap and that bound, which governs how much of a variance gain becomes earlier stopping; the descriptive HUNL EB-CS runs show a median $1.37\times$ stopping-time ratio. AV-AIVAT turns variance reduction into efficient, auditable early stopping while separating asymptotic screening from exact certification, so an evaluation can stop the moment its evidence suffices and hand a third party everything needed to recheck the verdict at that very stopping time.2. Resourced Authority A Mechanism-Design Model for Participatory Governance of Deployed AI Agents
- 来源：arXiv
- 日期：2026-08-06
- 作者/机构：Praphul Chandra, Sujit Gujar, Ganesh Ghalme
- 主题标签：
AI-Agent,arXiv - 推荐等级：高
- 分类：cs.GT, cs.AI, cs.MA
一句话结论
该机制设计通过资源分配实现AI代理的持续参与式治理，使授权自强制通过计算预算，建立Safe AI范式。
核心内容
- 机制基于资源分配原则，使AI代理授权自强制，建立Safe AI范式。
- 治理周期为序贯博弈，验证人类利益相关者依次在支持或拒绝市场贡献治理货币。
- 资金聚合器将贡献转化为广度加权支持，阈值门与耦合映射将支持转化为二元授权，释放计量计算预算。
方法与数据
- 方法为机制设计模型；数据摘要未明确。
价值判断
- 值得关注：提出计算作为治理杠杆的Safe AI范式，创新AI治理方法，为AI安全提供新视角。
- 可复用点：机制框架可复用于其他AI部署系统的参与式治理，无需额外基础设施。
- 局限/待核查：治理选民被代理操纵是核心开放问题，需进一步研究解决。