Image de couverture de livre

No Two Devils Alike: Unveiling Distinct Mechanisms of Fine-tuning Attacks

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Leong, Chak Tou, Cheng, Yi, Xu, Kaishuai, Wang, Jian, Wang, Hanlin, Li, Wenjie
Format:	Preprint
Publié:	2024
Sujets:	Computation and Language Cryptography and Security
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Why Safeguarded Ships Run Aground? Aligned Large Language Models' Safety Mechanisms Tend to Be Anchored in The Template Region
par: Leong, Chak Tou, et autres
Publié: (2025)

Mitigating Fine-tuning based Jailbreak Attack with Backdoor Enhanced Safety Alignment
par: Wang, Jiongxiao, et autres
Publié: (2024)

Practical Membership Inference Attacks against Fine-tuned Large Language Models via Self-prompt Calibration
par: Fu, Wenjie, et autres
Publié: (2023)

EnchTable: Unified Safety Alignment Transfer in Fine-tuned Large Language Models
par: Wu, Jialin, et autres
Publié: (2025)

LoRA-Leak: Membership Inference Attacks Against LoRA Fine-tuned Language Models
par: Ran, Delong, et autres
Publié: (2025)

Fun-tuning: Characterizing the Vulnerability of Proprietary LLMs to Optimization-based Prompt Injection Attacks via the Fine-Tuning Interface
par: Labunets, Andrey, et autres
Publié: (2025)

Window-based Membership Inference Attacks Against Fine-tuned Large Language Models
par: Chen, Yuetian, et autres
Publié: (2026)

E2CL: Exploration-based Error Correction Learning for Embodied Agents
par: Wang, Hanlin, et autres
Publié: (2024)

Safety is Not Only About Refusal: Reasoning-Enhanced Fine-tuning for Interpretable LLM Safety
par: Zhang, Yuyou, et autres
Publié: (2025)

Why LLM Safety Guardrails Collapse After Fine-tuning: A Similarity Analysis Between Alignment and Fine-tuning Datasets
par: Hsiung, Lei, et autres
Publié: (2025)

Seeing Isn't Believing: Mitigating Belief Inertia via Active Intervention in Embodied Agents
par: Wang, Hanlin, et autres
Publié: (2026)

STeCa: Step-level Trajectory Calibration for LLM Agent Learning
par: Wang, Hanlin, et autres
Publié: (2025)

Security Attacks on LLM-based Code Completion Tools
par: Cheng, Wen, et autres
Publié: (2024)

Token-level Data Selection for Safe LLM Fine-tuning
par: Li, Yanping, et autres
Publié: (2026)

SCOUT: A Defense Against Data Poisoning Attacks in Fine-Tuned Language Models
par: Afane, Mohamed, et autres
Publié: (2025)

Safety Anchor: Defending Harmful Fine-tuning via Geometric Bottlenecks
par: Lu, Guoxin, et autres
Publié: (2026)

Imitate Before Detect: Aligning Machine Stylistic Preference for Machine-Revised Text Detection
par: Chen, Jiaqi, et autres
Publié: (2024)

Evolve the Method, Not the Prompts: Evolutionary Synthesis of Jailbreak Attacks on LLMs
par: Chen, Yunhao, et autres
Publié: (2025)

MPMA: Preference Manipulation Attack Against Model Context Protocol
par: Wang, Zihan, et autres
Publié: (2025)

Virus: Harmful Fine-tuning Attack for Large Language Models Bypassing Guardrail Moderation
par: Huang, Tiansheng, et autres
Publié: (2025)

Beyond Jailbreaks: Revealing Stealthier and Broader LLM Security Risks Stemming from Alignment Failures
par: Zhou, Yukai, et autres
Publié: (2025)

Privately Learning from Graphs with Applications in Fine-tuning Large Language Models
par: Yin, Haoteng, et autres
Publié: (2024)

PRSA: Prompt Stealing Attacks against Real-World Prompt Services
par: Yang, Yong, et autres
Publié: (2024)

PandaGuard: Systematic Evaluation of LLM Safety against Jailbreaking Attacks
par: Shen, Guobin, et autres
Publié: (2025)

BadLingual: A Novel Lingual-Backdoor Attack against Large Language Models
par: Wang, Zihan, et autres
Publié: (2025)

A Quantitative Confirmation of the Currier Language Distinction
par: Parisel, Christophe
Publié: (2026)

ScaleOT: Privacy-utility-scalable Offsite-tuning with Dynamic LayerReplace and Selective Rank Compression
par: Yao, Kai, et autres
Publié: (2024)

SPA-RL: Reinforcing LLM Agents via Stepwise Progress Attribution
par: Wang, Hanlin, et autres
Publié: (2025)

Privacy in Large Language Models: Attacks, Defenses and Future Directions
par: Li, Haoran, et autres
Publié: (2023)

SEEP: Training Dynamics Grounds Latent Representation Search for Mitigating Backdoor Poisoning Attacks
par: He, Xuanli, et autres
Publié: (2024)

Private Fine-tuning of Large Language Models with Zeroth-order Optimization
par: Tang, Xinyu, et autres
Publié: (2024)

Efficient and Stealthy Jailbreak Attacks via Adversarial Prompt Distillation from LLMs to SLMs
par: Li, Xiang, et autres
Publié: (2025)

Separator Injection Attack: Uncovering Dialogue Biases in Large Language Models Caused by Role Separators
par: Li, Xitao, et autres
Publié: (2025)

TuBA: Cross-Lingual Transferability of Backdoor Attacks in LLMs with Instruction Tuning
par: He, Xuanli, et autres
Publié: (2024)

DualGuard: Dual-stream Large Language Model Watermarking Defense against Paraphrase and Spoofing Attack
par: Li, Hao, et autres
Publié: (2025)

Semantic-Preserving Adversarial Attacks on LLMs: An Adaptive Greedy Binary Search Approach
par: Zhang, Chong, et autres
Publié: (2025)

CTRAP: Embedding Collapse Trap to Safeguard Large Language Models from Harmful Fine-Tuning
par: Yi, Biao, et autres
Publié: (2025)

Black-Box Guardrail Reverse-engineering Attack
par: Yao, Hongwei, et autres
Publié: (2025)

BadApex: Backdoor Attack Based on Adaptive Optimization Mechanism of Black-box Large Language Models
par: Wu, Zhengxian, et autres
Publié: (2025)

Continuous Embedding Attacks via Clipped Inputs in Jailbreaking Large Language Models
par: Xu, Zihao, et autres
Publié: (2024)