惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

P
Proofpoint News Feed
V
V2EX
WordPress大学
WordPress大学
Google DeepMind News
Google DeepMind News
Martin Fowler
Martin Fowler
小众软件
小众软件
Blog — PlanetScale
Blog — PlanetScale
月光博客
月光博客
The Cloudflare Blog
T
Tailwind CSS Blog
H
Help Net Security
腾讯CDC
爱范儿
爱范儿
人人都是产品经理
人人都是产品经理
H
Hackread – Cybersecurity News, Data Breaches, AI and More
The GitHub Blog
The GitHub Blog
Microsoft Security Blog
Microsoft Security Blog
Stack Overflow Blog
Stack Overflow Blog
D
DataBreaches.Net
C
Check Point Blog
量子位
酷 壳 – CoolShell
酷 壳 – CoolShell
美团技术团队
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com

cs.CR updates on arXiv.org

On the Security of Research Artifacts SafeHarbor: Hierarchical Memory-Augmented Guardrail for LLM Agent Safety Agentic Vulnerability Reasoning on Windows COM Binaries From Beats to Breaches:How Offensive AI Infers Sensitive User Information from Playlists Undetectable Backdoors in Model Parameters: Hiding Sparse Secrets in High Dimensions When Embedding-Based Defenses Fail: Rethinking Safety in LLM-Based Multi-Agent Systems Token-Efficient Change Detection in LLM APIs Selfie-Capture Dynamics as an Auxiliary Signal Against Deepfakes and Injection Attacks for Mobile Identity Verification Trident: Improving Malware Detection with LLMs and Behavioral Features When Alignment Isn't Enough: Response-Path Attacks on LLM Agents RefusalGuard: Geometry-Preserving Fine-Tuning for Safety in LLMs Checkerboard: A Simple, Effective, Efficient and Learning-free Clean Label Backdoor Attack with Low Poisoning Budget Block-wise Codeword Embedding for Reliable Multi-bit Text Watermarking Secret Stealing Attacks on Local LLM Fine-Tuning through Supply-Chain Model Code Backdoors Enhancing Linux Privilege Escalation Attack Capabilities of Local LLM Agents Defusing the Trigger: Plug-and-Play Defense for Backdoored LLMs via Tail-Risk Intrinsic Geometric Smoothing Evaluating Jailbreaking Vulnerabilities in LLMs Deployed as Assistants for Smart Grid Operations: A Benchmark Against NERC Standards Behavioral Canaries: Auditing Private Retrieved Context Usage in RL Fine-Tuning FlexServe: A Fast and Secure LLM Serving System for Mobile Devices with Flexible Resource Isolation Breaking MCP with Function Hijacking Attacks: Novel Threats for Function Calling and Agentic Models Text Steganography with Dynamic Codebook and Multimodal Large Language Model TwoHamsters: Benchmarking Multi-Concept Compositional Unsafety in Text-to-Image Models Fundamental Limitations of Favorable Privacy-Utility Guarantees for DP-SGD Symbolic Guardrails for Domain-Specific Agents: Stronger Safety and Security Guarantees Without Sacrificing Utility Hardening x402: PII-Safe Agentic Payments via Pre-Execution Metadata Filtering QShield: Securing Neural Networks Against Adversarial Attacks using Quantum Circuits Hijacking Text Heritage: Hiding the Human Signature through Homoglyphic Substitution Like a Hammer, It Can Build, It Can Break: Large Language Model Uses, Perceptions, and Adoption in Cybersecurity Operations on Reddit Private Seeds, Public LLMs: Realistic and Privacy-Preserving Synthetic Data Generation One Word at a Time: Incremental Completion Decomposition Breaks LLM Safety
Patronus: Identifying and Mitigating Transferable Backdoo...
[Submitted on 7 Dec 2025 (v1), last revised 18 Jun 2026 (this ve · 2026-06-19 · via cs.CR updates on arXiv.org

View PDF HTML (experimental)

Abstract:The ``Pre-train, then fine-tune'' paradigm has revolutionized Natural Language Processing (NLP). In this context, transferable backdoors pose a severe threat to the Pre-trained Language Models (PLMs) supply chain, yet defensive research remains nascent, primarily relying on detecting anomalies in the output feature space. We identify a critical flaw that fine-tuning on downstream tasks inevitably modifies model parameters, shifting the output distribution and rendering pre-computed defense ineffective. To address this, we propose Patronus, a novel defense framework that shifts the defensive focus from output features to input-side invariance, exploiting the fact that adversarial triggers remain constant even as model weights change. To overcome the convergence challenges of discrete text optimization, Patronus introduces a multi-trigger contrastive search algorithm that effectively bridges gradient-based optimization with contrastive learning objectives. Furthermore, we employ a dual-stage mitigation strategy combining real-time input monitoring with model purification via adversarial training. Extensive experiments across 15 PLMs and nine tasks demonstrate that Patronus achieves $\geq98.3\%$ backdoor detection recall and reduces attack success rates to clean settings, significantly outperforming all state-of-the-art baselines in all settings. Code is available at this https URL.

Submission history

From: Haodong Zhao [view email]
[v1] Sun, 7 Dec 2025 15:51:56 UTC (1,319 KB)
[v2] Thu, 18 Jun 2026 07:18:15 UTC (1,512 KB)