Image de couverture de livre

Badllama 3: removing safety finetuning from Llama 3 in minutes

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteur principal:	Volkov, Dmitrii
Format:	Preprint
Publié:	2024
Sujets:	Machine Learning Artificial Intelligence 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

BadGPT-4o: stripping safety finetuning from GPT models
par: Krupkina, Ekaterina, et autres
Publié: (2024)

Emergent Misalignment: Narrow finetuning can produce broadly misaligned LLMs
par: Betley, Jan, et autres
Publié: (2025)

Forbidden Facts: An Investigation of Competing Objectives in Llama-2
par: Wang, Tony T., et autres
Publié: (2023)

Model Inversion Attacks on Llama 3: Extracting PII from Large Language Models
par: Sivashanmugam, Sathesh P.
Publié: (2025)

Llama-3.1-FoundationAI-SecurityLLM-Reasoning-8B Technical Report
par: Yang, Zhuoran, et autres
Publié: (2026)

LLM Agent Honeypot: Monitoring AI Hacking Agents in the Wild
par: Reworr, et autres
Publié: (2024)

Stealing User Prompts from Mixture of Experts
par: Yona, Itay, et autres
Publié: (2024)

Universal Jailbreak Backdoors from Poisoned Human Feedback
par: Rando, Javier, et autres
Publié: (2023)

Teach LLMs to Phish: Stealing Private Information from Language Models
par: Panda, Ashwinee, et autres
Publié: (2024)

Extracting Training Data from Diffusion Language Models via Infilling
par: Wang, Yihan, et autres
Publié: (2026)

There Are No Silly Questions: Evaluation of Offline LLM Capabilities from a Turkish Perspective
par: Yilmaz, Edibe, et autres
Publié: (2026)

Follow My Instruction and Spill the Beans: Scalable Data Extraction from Retrieval-Augmented Generation Systems
par: Qi, Zhenting, et autres
Publié: (2024)

Memories Retrieved from Many Paths: A Multi-Prefix Framework for Robust Detection of Training Data Leakage in Large Language Models
par: Dang, Trung Cuong, et autres
Publié: (2025)

AdvPrefix: An Objective for Nuanced LLM Jailbreaks
par: Zhu, Sicheng, et autres
Publié: (2024)

Best-of-Venom: Attacking RLHF by Injecting Poisoned Preference Data
par: Baumgärtner, Tim, et autres
Publié: (2024)

Covert Malicious Finetuning: Challenges in Safeguarding LLM Adaptation
par: Halawi, Danny, et autres
Publié: (2024)

TRAP: Targeted Random Adversarial Prompt Honeypot for Black-Box Identification
par: Gubri, Martin, et autres
Publié: (2024)

Fight Back Against Jailbreaking via Prompt Adversarial Tuning
par: Mo, Yichuan, et autres
Publié: (2024)

Gradient Cuff: Detecting Jailbreak Attacks on Large Language Models by Exploring Refusal Loss Landscapes
par: Hu, Xiaomeng, et autres
Publié: (2024)

Detecting Training Data of Large Language Models via Expectation Maximization
par: Kim, Gyuwan, et autres
Publié: (2024)

Machine Unlearning of Pre-trained Large Language Models
par: Yao, Jin, et autres
Publié: (2024)

Preserving Privacy in Large Language Models: A Survey on Current Threats and Solutions
par: Miranda, Michele, et autres
Publié: (2024)

DP-MemArc: Differential Privacy Transfer Learning for Memory Efficient Language Models
par: Liu, Yanming, et autres
Publié: (2024)

Towards Understanding the Fragility of Multilingual LLMs against Fine-Tuning Attacks
par: Poppi, Samuele, et autres
Publié: (2024)

Systematically Analyzing Prompt Injection Vulnerabilities in Diverse LLM Architectures
par: Benjamin, Victoria, et autres
Publié: (2024)

Rethinking How to Evaluate Language Model Jailbreak
par: Cai, Hongyu, et autres
Publié: (2024)

JailbreakRadar: Comprehensive Assessment of Jailbreak Attacks Against LLMs
par: Chu, Junjie, et autres
Publié: (2024)

Textual Unlearning Gives a False Sense of Unlearning
par: Du, Jiacheng, et autres
Publié: (2024)

Instructional Segment Embedding: Improving LLM Safety with Instruction Hierarchy
par: Wu, Tong, et autres
Publié: (2024)

SequentialBreak: Large Language Models Can be Fooled by Embedding Jailbreak Prompts into Sequential Prompt Chains
par: Saiem, Bijoy Ahmed, et autres
Publié: (2024)

LLMs can be Dangerous Reasoners: Analyzing-based Jailbreak Attack on Large Language Models
par: Lin, Shi, et autres
Publié: (2024)

Enhancing Prompt Injection Attacks to LLMs via Poisoning Alignment
par: Shao, Zedian, et autres
Publié: (2024)

SVIP: Towards Verifiable Inference of Open-source Large Language Models
par: Sun, Yifan, et autres
Publié: (2024)

Large Language Models in Cybersecurity: State-of-the-Art
par: Motlagh, Farzad Nourmohammadzadeh, et autres
Publié: (2024)

Towards Building a Robust Toxicity Predictor
par: Bespalov, Dmitriy, et autres
Publié: (2024)

Exploiting Class Probabilities for Black-box Sentence-level Attacks
par: Moraffah, Raha, et autres
Publié: (2024)

PostMark: A Robust Blackbox Watermark for Large Language Models
par: Chang, Yapei, et autres
Publié: (2024)

The Task Shield: Enforcing Task Alignment to Defend Against Indirect Prompt Injection in LLM Agents
par: Jia, Feiran, et autres
Publié: (2024)

BadAgent: Inserting and Activating Backdoor Attacks in LLM Agents
par: Wang, Yifei, et autres
Publié: (2024)

HSF: Defending against Jailbreak Attacks with Hidden State Filtering
par: Qian, Cheng, et autres
Publié: (2024)