paper-with-me

Papers

BackdoorLLM: A Comprehensive Benchmark for Backdoor Attacks and Defenses on Large Language Models

2024-08-23 · Yige Li, Hanxun Huang, Yunhan Zhao, Xingjun Ma, Jun Sun

Generative large language models (LLMs) have achieved state-of-the-art results on a wide range of tasks, yet they remain susceptible to backdoor attacks: carefully crafted triggers in the input can manipulate the model to produce adversary-specified outputs. While prior research has predominantly focused on backdoor risks in vision and classification settings, the vulnerability of LLMs in open-ended text generation remains underexplored. To fill this gap, we introduce BackdoorLLM (Our BackdoorLLM benchmark was awarded First Prize in the SafetyBench competition, https://www.mlsafety.org/safebench/winners, organized by the Center for AI Safety, https://safe.ai/.), the first comprehensive benchmark for systematically evaluating backdoor threats in text-generation LLMs. BackdoorLLM provides: (i) a unified repository of benchmarks with a standardized training and evaluation pipeline; (ii) a diverse suite of attack modalities, including data poisoning, weight poisoning, hidden-state manipulation, and chain-of-thought hijacking; (iii) over 200 experiments spanning 8 distinct attack strategies, 7 real-world scenarios, and 6 model architectures; (iv) key insights into the factors that govern backdoor effectiveness and failure modes in LLMs; and (v) a defense toolkit encompassing 7 representative mitigation techniques. Our code and datasets are available at https://github.com/bboylyg/BackdoorLLM. We will continuously incorporate emerging attack and defense methodologies to support the research in advancing the safety and reliability of LLMs.

📄 PDF Abstract BibTeX arXiv:2408.12798

Code (1)

bboylyg/backdoorllm 공식 구현 pytorch

Tasks

Data Poisoningtext-classificationText ClassificationText Generation

Similar Papers 제목 키워드 기반

Backdoor4Good: Benchmarking Beneficial Uses of Backdoors in LLMs

2026-03-08 · Yige Li, Wei Zhao, Zhe Li, Nay Myat Min 외 arxiv

Backdoor mechanisms have traditionally been studied as security threats that compromise the integrity of machine learning models. However, the same mechanism -- the conditional activation of specific behaviors through in…

AutoBackdoor: Automating Backdoor Attacks via LLM Agents

2025-11-20 · Yige Li, Zhe Li, Wei Zhao, Nay Myat Min 외 arxiv

Backdoor attacks pose a serious threat to the secure deployment of large language models (LLMs), enabling adversaries to implant hidden behaviors triggered by specific inputs. However, existing methods often rely on manu…

BackFed: An Efficient & Standardized Benchmark Suite for Backdoor Attacks in Federated Learning

2025-07-07 · Thinh Dao, Dung Thuy Nguyen, Khoa D Doan, Kok-Seng Wong

Federated Learning (FL) systems are vulnerable to backdoor attacks, where adversaries train their local models on poisoned data and submit poisoned model updates to compromise the global model. Despite numerous proposed …

Federated Learning

On Evaluating Neural Network Backdoor Defenses

2020-10-23 · Akshaj Veldanda, Siddharth Garg

Deep neural networks (DNNs) demonstrate superior performance in various fields, including scrutiny and security. However, recent studies have shown that DNNs are vulnerable to backdoor attacks. Several defenses were prop…

Backdoor Attacks and Defenses in Federated Learning: Survey, Challenges and Future Research Directions

2023-03-03 · Thuy Dung Nguyen, Tuan Nguyen, Phi Le Nguyen, Hieu H. Pham 외

Federated learning (FL) is a machine learning (ML) approach that allows the use of distributed data without compromising personal privacy. However, the heterogeneous distribution of data among clients in FL can make it d…

Backdoor AttackFederated LearningSurvey