发表机构
The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
ZoAQ通过重用历史查询实现自适应零阶优化，无需额外验证查询，在合成目标、黑盒攻击和OPT微调中分别减少43-48%查询、以低查询达到100%攻击成功率并节省43-46%前向评估。
AI 中文摘要
零阶优化（ZOO）通过函数评估来估计更新，使得扰动查询成为主要成本。固定预算在每一步使用相同数量的查询，而自适应控制器可能通过使用额外的预言机调用来测试估计器的可靠性，从而抵消其节省。我们引入了ZoAQ，一种围绕查询重用构建的自适应零阶优化方法。ZoAQ不会在每一步后丢弃过去的评估，而是使它们对下一次更新和决定是否进一步查询都有用。这使得无需额外验证查询即可实现自适应查询分配。我们的分析刻画了这种一致性何时能识别出支持下降的更新，并引导控制器达到足够的查询预算。在合成目标上，与使用120万次查询的固定基线相比，ZoAQ减少了43-48%的查询。在黑盒攻击中，它在MNIST和CIFAR-10上分别以平均320次和625次查询达到100%的成功率。在四种OPT微调设置中，相对于固定K=4，ZoAQ节省了43-46%的前向评估，任务内准确率变化范围从-0.018到+0.010。
英文摘要
Zeroth-order optimization (ZOO) estimates updates from function evaluations, making perturbation queries a primary cost. Fixed budgets spend the same number of queries at every step, while adaptive controllers may offset their savings by using additional oracle calls to test estimator reliability. We introduce ZoAQ, an adaptive ZOO method built around query reuse. Rather than discarding past evaluations after each step, ZoAQ makes them useful for both the next update and the decision to query further. This enables adaptive query allocation without extra validation queries. Our analysis characterizes when this agreement identifies an update that supports descent and guides the controller to a sufficient query budget. On synthetic objectives, ZoAQ reduces queries by 43-48% relative to fixed baselines using 1.2M queries. In black-box attacks, it reaches 100% success with 320 and 625 average queries on MNIST and CIFAR-10, respectively. Across four OPT fine-tuning settings, ZoAQ saves 43-46% forward evaluations relative to fixed K=4, with accuracy changes within tasks ranging from -0.018 to +0.010.
Comments40 pages, 13 figures, and 19 tables
