GradSentry: Gradient Spectral Entropy for Backdoor Sample Filtering in Large Language Model Fine-Tuning

Abstract

Fine-tuning Large Language Models with untrusted data exposes models to backdoor attacks, where poisoned samples cause targeted misbehavior. Existing sample-filtering defenses rely on clustering, which requires sufficient data and can fail at extreme poison ratios. We propose GradSentry ({Grad}ient {Sentry}), a backdoor sample filtering method based on the spectral entropy of per-sample gradients. Our key finding is that poisoned samples produce gradients with higher spectral entropy compared to

Categories

Framework mappings

Suggested from the entry's categories.

@misc{zhao2026gradsentry,
  title = {{GradSentry: Gradient Spectral Entropy for Backdoor Sample Filtering in Large Language Model Fine-Tuning}},
  author = {Haodong Zhao and Tianyi Xu and Tianhang Zhao and Zhuosheng Zhang and Gongshen Liu},
  year = {2026},
  month = may,
  eprint = {2605.26574},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2605.26574}
}