← Back to all news
arXiv AI By Zhaoxin Wang, Zihang Ding, Handing Wang

Sparse Fine-Tuning for Parameter-Efficient Adversarial Training

Read the original on arXiv AI →

The paper introduces Sparse Adversarial Fine‑Tuning (SAFT), a method that identifies a small set of Robustness‑Critical Parameters (RCPs) through adversarial gradient saliency and updates only these parameters during adversarial training. By focusing on this sparse subspace, SAFT achieves higher clean and robust accuracy than LoRA‑based parameter‑efficient methods and matches or surpasses full‑parameter adversarial training while training only about 5% of the parameters. The approach is validated across multiple architectures and datasets, demonstrating its effectiveness in reducing memory cost without sacrificing robustness.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv AI
Jul 9

LipSSD: Lipschitz-Constrained Single-Shot Detection for Adversarially Robust Object Detection

arXiv:2607. 06592v1 Announce Type: cross Abstract: Object detectors have many applications in safety-critical systems, but they are known to be sensitive to worst-case perturbations such as adversarial attacks, which limits their applicability in real-world scenarios.

By Vincent L\'eb\'e (IRIT, DTIPG - SNCF, UT3), Yannick Prudent (IRIT, DTIPG - SNCF, UT3), Corentin Friedrich (IRIT, DTIPG - SNCF, UT3), Thomas Massena (IRIT, DTIPG - SNCF, UT3), Ronan Sicre (IRIT), Franck Mamalet
computer-visionsafety
More like this →
arXiv Machine Learning
Oct 28

Augmentation-Guided Adversarial Training via Local Stability-Based Sample Selection

The paper introduces Augmentation-Guided Adversarial Training (AGAT), a selective adversarial training method that uses augmentation-induced local variation to identify unstable samples for training. AGAT employs Augmentation-based Local Stability (ALS) to rank samples based on the normalized change in model logits between an input and its label-preserving augmented version. Experiments on MNIST, CIFAR-10, CIFAR-100, and Tiny-ImageNet-200 show that AGAT achieves comparable or higher robust accuracy while applying adversarial supervision to only 20‑30% of training samples per minibatch, and it requires less training time than other adversarial training baselines.

By Daniel Agyei Asante, Mark Tetteh, Deborah Asamoah, Bernard Asante, Regina Esi Turkson
safety
More like this →