当隐私将机器学习中介的决策移至设备端：拍卖中的信息与激励错位
When Privacy Moves ML-Mediated Decisions On Device: Information and Incentive Misalignment in Auctions
查看机构详情
- Skelf Research（斯凯尔夫研究）
机构由 AI 辅助整理，请以论文原文为准。
浏览论文内容
中文总结 AI 辅助
本研究通过设备端拍卖模拟揭示隐私保护下机器学习决策的预算超支与激励错位问题，提出超额扣款界限并指出支付单位转换导致策略性偏离。
中文摘要 AI 辅助
将机器学习中介的决策移至保护隐私的客户端，使经济决策与推理一同去中心化。共享预算约束随后依赖于无法全局更新的信息，造成信息结构失效，而传统 pacing 方法并非为解决此问题而设计。我们在一个符合拍卖逻辑的设备端模拟中研究这种信息错位，模拟包含 36 个广告系列和 50 台设备。核算采用无量纲整数分数单位；不声称任何货币语义。在 30 条成对需求路径中，在原始 20 倍预算压力下，比例偶数 pacing 在一次滞后 tick 后超支 17.77%，在 50 个 tick 后超支 1,669.31%。该效应并不依赖于如此严重的预算：在 2 倍压力下，50 tick 超支仍为 106.95%。一个可见预算的无销售保护在零滞后时在该分数单位粒度下实现精确合规，但在一个 tick 时仍留下 11.88% 的超支，因为其他设备的扣款仍不可见。一个声明的突发、异构设备扫描保持严格递增的平均滞后曲线。我们在条件收费上限下推导出有限窗口的预期超额扣款界限，并在每个有界值单元中发现正的成对松弛。当允许 ML/pacing 分数转换改变支付单位时，出现第二种激励错位：在一个 tick 时，98.23% 的竞争拍卖允许有利可图的偏离。一个可执行的实现级反例将亚军的乘数隔离在赢家价格中。关键基础出价支付在给定当前乘数下是每次拍卖的 DSIC，但并未建立动态真实性，也未修复基础价值排名分歧。
英文摘要
Moving ML-mediated decision making onto privacy-preserving clients decentralises the economic decision along with the inference. Shared budget constraints then depend on information that cannot be globally current, creating an information-structure failure that conventional pacing is not designed to solve. We study this information misalignment in an auction-logic-faithful on-device simulation with 36 campaigns and 50 devices. Accounting is in dimensionless integer score units; no currency semantics are claimed. Across 30 paired demand paths, proportional Even pacing overspends 17.77% after one tick of staleness and 1,669.31% after 50 ticks under the original 20-times budget pressure. The effect does not depend on that severe a budget: at two-times pressure, 50-tick overspend remains 106.95%. A visible-budget no-sale guard makes zero-lag compliance exact at this score-unit granularity, yet leaves 11.88% overspend at one tick because other devices' debits remain invisible. A declared bursty, heterogeneous-device sweep retains a strictly increasing mean lag curve. We derive a finite-window expected excess-debit bound under conditional charge caps and find positive paired slack in every bounded-value cell. A second, incentive misalignment arises when the ML/pacing score transformation is allowed to change payment units: 98.23% of rival auctions at one tick admit a profitable deviation. An executable implementation-level counterexample isolates the runner-up's multiplier in the winner's price. Critical-base-bid payment is per-auction DSIC conditional on current multipliers, but does not establish dynamic truthfulness and does not repair base-value ranking disagreement.
