Pith. sign in

Sparse-RS: a versatile framework for query-efficient sparse black-box adversarial attacks

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
abstract

We propose a versatile framework based on random search, Sparse-RS, for score-based sparse targeted and untargeted attacks in the black-box setting. Sparse-RS does not rely on substitute models and achieves state-of-the-art success rate and query efficiency for multiple sparse attack models: $l_0$-bounded perturbations, adversarial patches, and adversarial frames. The $l_0$-version of untargeted Sparse-RS outperforms all black-box and even all white-box attacks for different models on MNIST, CIFAR-10, and ImageNet. Moreover, our untargeted Sparse-RS achieves very high success rates even for the challenging settings of $20\times20$ adversarial patches and $2$-pixel wide adversarial frames for $224\times224$ images. Finally, we show that Sparse-RS can be applied to generate targeted universal adversarial patches where it significantly outperforms the existing approaches. The code of our framework is available at https://github.com/fra31/sparse-rs.

fields

cs.LG 1

years

2026 1

verdicts

ACCEPT 1

representative citing papers

LaPrune: Controllable Differentiable Sparsity at Million Scale

cs.LG · 2026-08-04 · accept · novelty 6.0

A differentiable top-k mask layer that enforces an exact selection budget and uses a normalized hardness parameter to interpolate from equal-weight masks to hard binary masks, with saturation theory and million-scale results.

citing papers explorer

Showing 1 of 1 citing paper.

  • LaPrune: Controllable Differentiable Sparsity at Million Scale cs.LG · 2026-08-04 · accept · none · ref 40 · internal anchor

    A differentiable top-k mask layer that enforces an exact selection budget and uses a normalized hardness parameter to interpolate from equal-weight masks to hard binary masks, with saturation theory and million-scale results.