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

L
LangChain Blog
阮一峰的网络日志
阮一峰的网络日志
WordPress大学
WordPress大学
博客园 - 司徒正美
罗磊的独立博客
D
Docker
Last Week in AI
Last Week in AI
爱范儿
爱范儿
M
MIT News - Artificial intelligence
V
V2EX
Google DeepMind News
Google DeepMind News
小众软件
小众软件
Apple Machine Learning Research
Apple Machine Learning Research
Microsoft Security Blog
Microsoft Security Blog
T
Tailwind CSS Blog
MyScale Blog
MyScale Blog
V
Visual Studio Blog
博客园 - 叶小钗
B
Blog RSS Feed
A
About on SuperTechFans
F
Fortinet All Blogs
T
The Blog of Author Tim Ferriss
Martin Fowler
Martin Fowler
P
Proofpoint News Feed

cs.CR updates on arXiv.org

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 An AI Agent Execution Environment to Safeguard User Data 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 Measuring and Exploiting Contextual Bias in LLM-Assisted Security Code Review
Language Models May Verbatim Complete Text They Were Not ...
Ken Ziyu Liu, Christopher A. Choquette-Choo, Matthew Jagielski, · 2025-03-22 · via cs.CR updates on arXiv.org

An important question today is whether a given text was used to train a large language model (LLM). A \emph{completion} test is often employed: check if the LLM completes a sufficiently complex text. This, however, requires a ground-truth definition of membership; most commonly, it is defined as a member based on the $n$-gram overlap between the target text and any text in the dataset. In this work, we demonstrate that this $n$-gram based membership definition can be effectively gamed. We study scenarios where sequences are \emph{non-members} for a given $n$ and we find that completion tests still succeed. We find many natural cases of this phenomenon by retraining LLMs from scratch after removing all training samples that were completed; these cases include exact duplicates, near-duplicates, and even short overlaps. They showcase that it is difficult to find a single viable choice of $n$ for membership definitions. Using these insights, we design adversarial datasets that can cause a given target sequence to be completed without containing it, for any reasonable choice of $n$. Our findings highlight the inadequacy of $n$-gram membership, suggesting membership definitions fail to account for auxiliary information available to the training algorithm.