Breaking the Loop: Detecting and Mitigating Denial-of-Service Vulnerabilities in Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yu, Junzhe, Liu, Yi, Sun, Huijia, Shi, Ling, Chen, Yuqi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912254951161856
author Yu, Junzhe
Liu, Yi
Sun, Huijia
Shi, Ling
Chen, Yuqi
author_facet Yu, Junzhe
Liu, Yi
Sun, Huijia
Shi, Ling
Chen, Yuqi
contents Large Language Models (LLMs) have significantly advanced text understanding and generation, becoming integral to applications across education, software development, healthcare, entertainment, and legal services. Despite considerable progress in improving model reliability, latency remains under-explored, particularly through recurrent generation, where models repeatedly produce similar or identical outputs, causing increased latency and potential Denial-of-Service (DoS) vulnerabilities. We propose RecurrentGenerator, a black-box evolutionary algorithm that efficiently identifies recurrent generation scenarios in prominent LLMs like LLama-3 and GPT-4o. Additionally, we introduce RecurrentDetector, a lightweight real-time classifier trained on activation patterns, achieving 95.24% accuracy and an F1 score of 0.87 in detecting recurrent loops. Our methods provide practical solutions to mitigate latency-related vulnerabilities, and we publicly share our tools and data to support further research.
format Preprint
id arxiv_https___arxiv_org_abs_2503_00416
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Breaking the Loop: Detecting and Mitigating Denial-of-Service Vulnerabilities in Large Language Models
Yu, Junzhe
Liu, Yi
Sun, Huijia
Shi, Ling
Chen, Yuqi
Cryptography and Security
Artificial Intelligence
Performance
Large Language Models (LLMs) have significantly advanced text understanding and generation, becoming integral to applications across education, software development, healthcare, entertainment, and legal services. Despite considerable progress in improving model reliability, latency remains under-explored, particularly through recurrent generation, where models repeatedly produce similar or identical outputs, causing increased latency and potential Denial-of-Service (DoS) vulnerabilities. We propose RecurrentGenerator, a black-box evolutionary algorithm that efficiently identifies recurrent generation scenarios in prominent LLMs like LLama-3 and GPT-4o. Additionally, we introduce RecurrentDetector, a lightweight real-time classifier trained on activation patterns, achieving 95.24% accuracy and an F1 score of 0.87 in detecting recurrent loops. Our methods provide practical solutions to mitigate latency-related vulnerabilities, and we publicly share our tools and data to support further research.
title Breaking the Loop: Detecting and Mitigating Denial-of-Service Vulnerabilities in Large Language Models
topic Cryptography and Security
Artificial Intelligence
Performance
url https://arxiv.org/abs/2503.00416