paper-with-me

Papers

Efficient but Vulnerable: Benchmarking and Defending LLM Batch Prompting Attack

2025-03-18 · Murong Yue, Ziyu Yao

Batch prompting, which combines a batch of multiple queries sharing the same context in one inference, has emerged as a promising solution to reduce inference costs. However, our study reveals a significant security vulnerability in batch prompting: malicious users can inject attack instructions into a batch, leading to unwanted interference across all queries, which can result in the inclusion of harmful content, such as phishing links, or the disruption of logical reasoning. In this paper, we construct BATCHSAFEBENCH, a comprehensive benchmark comprising 150 attack instructions of two types and 8k batch instances, to study the batch prompting vulnerability systematically. Our evaluation of both closed-source and open-weight LLMs demonstrates that all LLMs are susceptible to batch-prompting attacks. We then explore multiple defending approaches. While the prompting-based defense shows limited effectiveness for smaller LLMs, the probing-based approach achieves about 95% accuracy in detecting attacks. Additionally, we perform a mechanistic analysis to understand the attack and identify attention heads that are responsible for it.

📄 PDF Abstract BibTeX arXiv:2503.15551

Code (0)

등록된 구현이 없습니다.

Tasks

8kBenchmarkingLogical Reasoning

Methods 이 논문이 사용한 방법론

Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…
Attention 설명 없음

Similar Papers 제목 키워드 기반

Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming

2025-01-31 · Mrinank Sharma, Meg Tong, Jesse Mu, Jerry Wei 외

Large language models (LLMs) are vulnerable to universal jailbreaks-prompting strategies that systematically bypass model safeguards and enable users to carry out harmful processes that require many model interactions, l…

Red Teaming

Universal Adversarial Training with Class-Wise Perturbations

2021-04-07 · Philipp Benz, Chaoning Zhang, Adil Karjauv, In So Kweon

Despite their overwhelming success on a wide range of applications, convolutional neural networks (CNNs) are widely recognized to be vulnerable to adversarial examples. This intriguing phenomenon led to a competition bet…

Adversarial Robustness

Towards Defending Multiple $\ell_p$-norm Bounded Adversarial Perturbations via Gated Batch Normalization

2020-12-03 · Aishan Liu, Shiyu Tang, Xinyun Chen, Lei Huang 외

There has been extensive evidence demonstrating that deep neural networks are vulnerable to adversarial examples, which motivates the development of defenses against adversarial attacks. Existing adversarial defenses typ…

GenTel-Safe: A Unified Benchmark and Shielding Framework for Defending Against Prompt Injection Attacks

2024-09-29 · Rongchang Li, Minjie Chen, Chang Hu, Han Chen 외

Large Language Models (LLMs) like GPT-4, LLaMA, and Qwen have demonstrated remarkable success across a wide range of applications. However, these models remain inherently vulnerable to prompt injection attacks, which can…

Benchmarking

MultiRobustBench: Benchmarking Robustness Against Multiple Attacks

2023-02-21 · Sihui Dai, Saeed Mahloujifar, Chong Xiang, Vikash Sehwag 외

The bulk of existing research in defending against adversarial examples focuses on defending against a single (typically bounded Lp-norm) attack, but for a practical setting, machine learning (ML) models should be robust…

Benchmarking