锚定对抗轨迹到数据流形：一种双层迁移优化框架
Anchoring Adversarial Trajectories to Data Manifolds: A Bilevel Transfer Optimization Framework
- The University of Hong Kong（香港大学）
- The Hong Kong Polytechnic University（香港理工大学）
机构由 AI 辅助整理，请以论文原文为准。
中文总结 AI 辅助
针对对抗迁移中轨迹偏离数据流形的问题，提出MABT框架，通过流形锚定与双层优化提升10个基线攻击者在28种配置下的迁移性。
中文摘要 AI 辅助
对抗迁移的一个关键瓶颈是轨迹层面的几何脱节：环境梯度常常偏离内在数据流形，导致代理特定过拟合。为解决此问题，我们提出流形锚定双层迁移（MABT），一个将对抗轨迹锚定到共享语义子空间的统一框架。MABT引入一个松弛的流形锚定算子作为语义修正器，以抑制离流形噪声。在此约束下，我们将迁移攻击生成视为一个分布双层优化问题，通过最小化代理不确定性分布下的期望迁移风险来学习几何对齐的初始化。我们进一步开发了一个具有线性时间复杂度的免Hessian求解器来处理由此产生的层级结构。实验表明，在28种攻击配置、多种受害者架构和防御机制下，10个基线攻击者的迁移性得到提升。
英文摘要
A key bottleneck in adversarial transfer is a trajectory-level geometric disconnect: ambient gradients often drift away from the intrinsic data manifold, causing surrogate-specific overfitting. To rectify this, we propose Manifold Anchored Bilevel Transfer (MABT), a unified framework that anchors adversarial trajectories to the shared semantic subspace. MABT introduces a relaxed manifold-anchoring operator as a semantic rectifier to suppress off-manifold noise. With this constraint, we cast transfer attack generation as a distributional bilevel optimization problem that learns a geometry-aligned initialization by minimizing expected transfer risk under a surrogate uncertainty distribution. We further develop a Hessian-free solver with linear-time complexity to handle the resulting hierarchy. Experiments demonstrate improved transferability for 10 baseline attackers across 28 attack configurations, diverse victim architectures, and defense mechanisms.
