cs.LGNov 5, 2026

RLGPA: A Reinforcement Learning-Guided Granularity Planning Attack for Decision-Based Sparse Adversarial Examples

Authors: Chao Zhou, Shenghua He, Jiaxin Mo, Qiangguo Yu, Fengchuang Xing

Organizations: School of Electronic Information, Huzhou College, Huzhou, 313000, China · School of Electronic Science and Engineering, South China Normal University, Foshan, 528225, China · School of Physics and Information Engineering, Guangdong University of Education, Guangzhou, 510303, China

Abstract

Decision-based sparse attacks must reduce perturbation support using only the victim's predicted label. The best restoration strategy varies as support contracts: coarse removal works early, while tail states need finer, history-aware tests. We propose RLGPA, an online reinforcement learning method that schedules support-reduction executors and their granularity under a hard query budget. A masked Dueling Double DQN selects among seven stable policy--scale actions; each action produces support-relative restoration groups that are rebuilt after every macro decision. A feasibility-preserving executor commits only adversarial restorations and feeds support gain and query cost back to the planner. Under the matched top-1-label oracle, RLGPA reaches 61% targeted ASR at sparsity 0.01 on CIFAR-10 (vs. 25% SparseEvo, 0% Pointwise), and 75.5%/73.0% on ResNet50/ViT at sparsity 0.005 (vs. 29.0%/15.5%). A fixed-probe test reduces median L0L_0 from 165 to 23 after 40 images without any update. State-blind and state-shuffled controls degrade significantly, confirming that the learned policy uses attack state; the main gains come from state-conditioned executor scheduling and dynamic repartition. The source code is publicly available at https://github.com/Acui001/RLGPA.git.

Figures & tables

Explore similar work

CardsList
  1. Sparse Fine-Tuning for Parameter-Efficient Adversarial Training

    Oct 29, 2026Zhaoxin Wang, Zihang Ding, Handing WangAdversarial TrainingAdversarial Robustness

  2. Latent Geometric Chords for Query-Efficient Decision-Based Adversarial Attacks

    May 29, 2026Ei Hmue Khine, Yao Li, Jiebao Sun +3Adversarial AttacksAdversarial Perturbations