Cover Image

Toward Principled LLM Safety Testing: Solving the Jailbreak Oracle Problem

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Lin, Shuyi, Suri, Anshuman, Oprea, Alina, Tan, Cheng
Format:	Preprint
Published:	2025
Subjects:	Cryptography and Security Artificial Intelligence Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Toward a Principled Framework for Agent Safety Measurement
by: Lin, Shuyi, et al.
Published: (2026)

SAGA: A Security Architecture for Governing AI Agentic Systems
by: Syros, Georgios, et al.
Published: (2025)

Syntax- and Compilation-Preserving Evasion of LLM Vulnerability Detectors
by: Sun, Luze, et al.
Published: (2026)

Riddle Me This! Stealthy Membership Inference for Retrieval-Augmented Generation
by: Naseh, Ali, et al.
Published: (2025)

Do Parameters Reveal More than Loss for Membership Inference?
by: Suri, Anshuman, et al.
Published: (2024)

Dissecting Distribution Inference
by: Suri, Anshuman, et al.
Published: (2022)

Text-to-Image Models Leave Identifiable Signatures: Implications for Leaderboard Security
by: Naseh, Ali, et al.
Published: (2025)

Exploiting Leaderboards for Large-Scale Distribution of Malicious Models
by: Suri, Anshuman, et al.
Published: (2025)

PoolFlip: A Multi-Agent Reinforcement Learning Security Environment for Cyber Defense
by: Cadet, Xavier, et al.
Published: (2025)

The Art of the Jailbreak: Formulating Jailbreak Attacks for LLM Security Beyond Binary Scoring
by: Hossain, Ismail, et al.
Published: (2026)

Odysseus: Jailbreaking Commercial Multimodal LLM-integrated Systems via Dual Steganography
by: Li, Songze, et al.
Published: (2025)

Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks
by: Andriushchenko, Maksym, et al.
Published: (2024)

Fuzz-Testing Meets LLM-Based Agents: An Automated and Efficient Framework for Jailbreaking Text-To-Image Generation Models
by: Dong, Yingkai, et al.
Published: (2024)

Bypassing LLM Guardrails: An Empirical Analysis of Evasion Attacks against Prompt Injection and Jailbreak Detection Systems
by: Hackett, William, et al.
Published: (2025)

DROP: Poison Dilution via Knowledge Distillation for Federated Learning
by: Syros, Georgios, et al.
Published: (2025)

Immune: Improving Safety Against Jailbreaks in Multi-modal LLMs via Inference-Time Alignment
by: Ghosal, Soumya Suvra, et al.
Published: (2024)

The Jailbreak Tax: How Useful are Your Jailbreak Outputs?
by: Nikolić, Kristina, et al.
Published: (2025)

Identifying Models Behind Text-to-Image Leaderboards
by: Naseh, Ali, et al.
Published: (2026)

Functional Homotopy: Smoothing Discrete Optimization via Continuous Parameters for LLM Jailbreak Attacks
by: Wang, Zi, et al.
Published: (2024)

Mitigating Many-Shot Jailbreaking
by: Ackerman, Christopher M., et al.
Published: (2025)

AdvPrefix: An Objective for Nuanced LLM Jailbreaks
by: Zhu, Sicheng, et al.
Published: (2024)

Uncovering Logit Suppression Vulnerabilities in LLM Safety Alignment
by: Li, Yuxi, et al.
Published: (2024)

Injecting Universal Jailbreak Backdoors into LLMs in Minutes
by: Chen, Zhuowei, et al.
Published: (2025)

Jailbreaking and Mitigation of Vulnerabilities in Large Language Models
by: Peng, Benji, et al.
Published: (2024)

Scalable Defense against In-the-wild Jailbreaking Attacks with Safety Context Retrieval
by: Chen, Taiye, et al.
Published: (2025)

On the (In)feasibility of ML Backdoor Detection as an Hypothesis Testing Problem
by: Pichler, Georg, et al.
Published: (2024)

Attention Eclipse: Manipulating Attention to Bypass LLM Safety-Alignment
by: Zaree, Pedram, et al.
Published: (2025)

A Causal Perspective for Enhancing Jailbreak Attack and Defense
by: Pan, Licheng, et al.
Published: (2026)

GAVEL: Towards Rule-Based Safety Through Activation Monitoring
by: Rozenfeld, Shir, et al.
Published: (2026)

Jailbreaking LLMs Without Gradients or Priors: Effective and Transferable Attacks
by: Nurlanov, Zhakshylyk, et al.
Published: (2026)

SoK: Pitfalls in Evaluating Black-Box Attacks
by: Suya, Fnu, et al.
Published: (2023)

Not All Tokens Are Created Equal: Query-Efficient Jailbreak Fuzzing for LLMs
by: Chen, Wenyu, et al.
Published: (2026)

Improved Large Language Model Jailbreak Detection via Pretrained Embeddings
by: Galinkin, Erick, et al.
Published: (2024)

The Erasure Illusion: Stress-Testing the Generalization of LLM Forgetting Evaluation
by: Jia, Hengrui, et al.
Published: (2025)

Fewer Weights, More Problems: A Practical Attack on LLM Pruning
by: Egashira, Kazuki, et al.
Published: (2025)

Jailbreaking GPT-4V via Self-Adversarial Attacks with System Prompts
by: Wu, Yuanwei, et al.
Published: (2023)

GuardReasoner: Towards Reasoning-based LLM Safeguards
by: Liu, Yue, et al.
Published: (2025)

Unsafe LLM-Based Search: Quantitative Analysis and Mitigation of Safety Risks in AI Web Search
by: Luo, Zeren, et al.
Published: (2025)

Guiding not Forcing: Enhancing the Transferability of Jailbreaking Attacks on LLMs via Removing Superfluous Constraints
by: Yang, Junxiao, et al.
Published: (2025)

AutoDAN-Turbo: A Lifelong Agent for Strategy Self-Exploration to Jailbreak LLMs
by: Liu, Xiaogeng, et al.
Published: (2024)