GradSentry: Gradient Spectral Entropy for Backdoor Sample Filtering in Large Language Model Fine-Tuning
:::info Stub — Full Engineering Breakdown Coming This paper was featured on Hugging Face Daily Papers on 2026-05-26 with 10 upvotes. A full breakdown with production viability rating, implementation notes, and honest limitations is being written. Subscribe to AI Letters → :::
| Authors | Haodong Zhao et al. |
| Year | 2026 |
| HF Upvotes | 10 |
| arXiv | 2605.26574 |
| Download | |
| HF Page | View on Hugging Face |
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 clean samples. GradSentry captures output-altering backdoor signatures using per-sample gradient spectra, avoiding pairwise sample comparisons and clustering during feature construction. Importantly, our method is training-agnostic: it works for both parameter-efficient fine-tuning methods like LoRA and full-parameter tuning, as the gradient analysis operates independently of which parameters are being updated during training. GradSentry requires no clustering, operates effectively across all poison ratios (1%--90%), and introduces minimal computational overhead (20-50ms per sample for 7B model). Evaluation on four QA datasets and four attack types demonstrates the effectiveness of spectral entropy for backdoor detection. Code is available at https://github.com/dongdongzhaoUP/GradSentry.
Engineering Breakdown
The Problem
Existing sample-filtering defenses rely on clustering, which requires sufficient data and can fail at extreme poison ratios.
The Approach
We propose GradSentry ({Grad}ient {Sentry}), a backdoor sample filtering method based on the spectral entropy of per-sample gradients. Importantly, our method is training-agnostic: it works for both parameter-efficient fine-tuning methods like LoRA and full-parameter tuning, as the gradient analysis operates independently of which parameters are being updated during training.
Key Results
Code is available at https://github.com/dongdongzhaoUP/GradSentry.
Research Areas
This paper contributes to the following areas of AI/ML engineering:
- Machine learning
- Deep learning
- Neural networks
- Model optimization
- AI systems
- Gradsentry
:::tip Subscribe Get weekly breakdowns of papers like this in AI Letters - the newsletter for engineers building production AI systems. :::