Harmful Fine-tuning Attacks and Defenses for Large Language Models: A Survey

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Huang, Tiansheng, Hu, Sihao, Ilhan, Fatih, Tekin, Selim Furkan, Liu, Ling
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913058066006016
author Huang, Tiansheng
Hu, Sihao
Ilhan, Fatih
Tekin, Selim Furkan
Liu, Ling
author_facet Huang, Tiansheng
Hu, Sihao
Ilhan, Fatih
Tekin, Selim Furkan
Liu, Ling
contents Recent research demonstrates that the nascent fine-tuning-as-a-service business model exposes serious safety concerns: fine-tuning with a few harmful data uploaded from the users can compromise the safety alignment of the model. The attack, known as harmful fine-tuning attack, has generated broad research interests in both academia and industry. In this paper, we first systematically formulate the threat model and basic assumptions of harmful fine-tuning. Then, we provide a comprehensive review of harmful fine-tuning from three fundamental perspectives: attack setting, defense design, and evaluation methodology. First, we present the threat model of the problem and introduce the harmful fine-tuning attack and its variants. Next, we systematically survey representative attacks, defense methods, and mechanical analysis of adverse effects in the existing literature. Finally, we introduce the evaluation methodology and outline future research directions, which can serve as guidelines and crucial perspectives for the future development of the subject. We also maintain a curated list of relevant papers, which are made accessible at https://github.com/git-disl/awesome_LLM-harmful-fine-tuning-papers
format Preprint
id arxiv_https___arxiv_org_abs_2409_18169
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Harmful Fine-tuning Attacks and Defenses for Large Language Models: A Survey
Huang, Tiansheng
Hu, Sihao
Ilhan, Fatih
Tekin, Selim Furkan
Liu, Ling
Cryptography and Security
Artificial Intelligence
Machine Learning
Recent research demonstrates that the nascent fine-tuning-as-a-service business model exposes serious safety concerns: fine-tuning with a few harmful data uploaded from the users can compromise the safety alignment of the model. The attack, known as harmful fine-tuning attack, has generated broad research interests in both academia and industry. In this paper, we first systematically formulate the threat model and basic assumptions of harmful fine-tuning. Then, we provide a comprehensive review of harmful fine-tuning from three fundamental perspectives: attack setting, defense design, and evaluation methodology. First, we present the threat model of the problem and introduce the harmful fine-tuning attack and its variants. Next, we systematically survey representative attacks, defense methods, and mechanical analysis of adverse effects in the existing literature. Finally, we introduce the evaluation methodology and outline future research directions, which can serve as guidelines and crucial perspectives for the future development of the subject. We also maintain a curated list of relevant papers, which are made accessible at https://github.com/git-disl/awesome_LLM-harmful-fine-tuning-papers
title Harmful Fine-tuning Attacks and Defenses for Large Language Models: A Survey
topic Cryptography and Security
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2409.18169