# What Features in Prompts Jailbreak LLMs? Investigating the Mechanisms Behind Attacks

arXiv: https://arxiv.org/abs/2411.03343 · HTML: https://mlsanity.com/paper/2411.03343/

## Key facts

- Cited by 29 papers in the map (data through 14 Oct 2026).
- Placed in Jailbreaking, one of 12 directions in LLM safety and security.
- Builds on 30 of its 47 references that are in the map.
- First posted on arXiv 2 Nov 2024.

## Placed in

- LLM safety and security / [Jailbreaking](https://mlsanity.com/safety/jailbreaking/)
- Mechanistic interpretability / [Steering and probing for LLM safety](https://mlsanity.com/mechinterp/steering-and-probing-for-llm-safety/)

## Builds on (most cited first, 30 of 30)

- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [Jailbroken: How Does LLM Safety Training Fail?](https://mlsanity.com/paper/2307.02483/) (2023)
- [HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal](https://mlsanity.com/paper/2402.04249/) (2024)
- [Understanding intermediate layers using linear classifier probes](https://mlsanity.com/paper/1610.01644/) (2016)
- [Representation Engineering: A Top-Down Approach to AI Transparency](https://mlsanity.com/paper/2310.01405/) (2023)
- [Inference-Time Intervention: Eliciting Truthful Answers from a Language Model](https://mlsanity.com/paper/2306.03341/) (2023)
- [Steering Llama 2 via Contrastive Activation Addition](https://mlsanity.com/paper/2312.06681/) (2023)
- [Refusal in Language Models Is Mediated by a Single Direction](https://mlsanity.com/paper/2406.11717/) (2024)
- [Probing Classifiers: Promises, Shortcomings, and Advances](https://mlsanity.com/paper/2102.12452/) (2021)
- [AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models](https://mlsanity.com/paper/2310.04451/) (2023)
- [Discovering Latent Knowledge in Language Models Without Supervision](https://mlsanity.com/paper/2212.03827/) (2022)
- [The Linear Representation Hypothesis and the Geometry of Large Language Models](https://mlsanity.com/paper/2311.03658/) (2023)
- ["Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models](https://mlsanity.com/paper/2308.03825/) (2023)
- [What you can cram into a single vector: Probing sentence embeddings for linguistic properties](https://mlsanity.com/paper/1805.01070/) (2018)
- [A StrongREJECT for Empty Jailbreaks](https://mlsanity.com/paper/2402.10260/) (2024)
- [Jailbreak Attacks and Defenses Against Large Language Models: A Survey](https://mlsanity.com/paper/2407.04295/) (2024)
- [WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Models](https://mlsanity.com/paper/2406.18510/) (2024)
- [Foundational Challenges in Assuring Alignment and Safety of Large Language Models](https://mlsanity.com/paper/2404.09932/) (2024)
- [Survey of Vulnerabilities in Large Language Models Revealed by Adversarial Attacks](https://mlsanity.com/paper/2310.10844/) (2023)
- [Latent Adversarial Training Improves Robustness to Persistent Harmful Behaviors in LLMs](https://mlsanity.com/paper/2407.15549/) (2024)
- [Probing the Probing Paradigm: Does Probing Accuracy Entail Task Relevance?](https://mlsanity.com/paper/2005.00719/) (2020)
- [Probing Pre-Trained Language Models for Cross-Cultural Differences in Values](https://mlsanity.com/paper/2203.13722/) (2022)
- [On Prompt-Driven Safeguarding for Large Language Models](https://mlsanity.com/paper/2401.18018/) (2024)
- [Tricking LLMs into Disobedience: Formalizing, Analyzing, and Detecting Jailbreaks](https://mlsanity.com/paper/2305.14965/) (2023)
- [Steering Without Side Effects: Improving Post-Deployment Control of Language Models](https://mlsanity.com/paper/2406.15518/) (2024)
- [Breaking Down the Defenses: A Comparative Survey of Attacks on Large Language Models](https://mlsanity.com/paper/2403.04786/) (2024)
- [Understanding Jailbreak Success: A Study of Latent Space Dynamics in Large Language Models](https://mlsanity.com/paper/2406.09289/) (2024)
- [Social Bias Probing: Fairness Benchmarking for Language Models](https://mlsanity.com/paper/2311.09090/) (2023)
- [Towards Inference-time Category-wise Safety Steering for Large Language Models](https://mlsanity.com/paper/2410.01174/) (2024)

## Built on by (most cited first, 29 of 29)

- [RL-Obfuscation: Can Language Models Learn to Evade Latent-Space Monitors?](https://mlsanity.com/paper/2506.14261/) (2025)
- [PIShield: Detecting Prompt Injection Attacks via Intrinsic LLM Features](https://mlsanity.com/paper/2510.14005/) (2025)
- [Understanding and Mitigating Overrefusal in LLMs from an Unveiling Perspective of Safety Decision Boundary](https://mlsanity.com/paper/2505.18325/) (2025)
- [Refusal Behavior in Large Language Models: A Nonlinear Perspective](https://mlsanity.com/paper/2501.08145/) (2025)
- [Universal Jailbreak Suffixes Are Strong Attention Hijackers](https://mlsanity.com/paper/2506.12880/) (2025)
- [The Impact of Off-Policy Training Data on Probe Generalisation](https://mlsanity.com/paper/2511.17408/) (2025)
- [Why Safeguarded Ships Run Aground? Aligned Large Language Models' Safety Mechanisms Tend to Be Anchored in The Template Region](https://mlsanity.com/paper/2502.13946/) (2025)
- [DUAL-Bench: Measuring Over-Refusal and Robustness in Vision-Language Models](https://mlsanity.com/paper/2510.10846/) (2025)
- [PII Jailbreaking in LLMs via Activation Steering Reveals Personal Information Leakage](https://mlsanity.com/paper/2507.02332/) (2025)
- [Your Agent is More Brittle Than You Think: Uncovering Indirect Injection Vulnerabilities in Agentic LLMs](https://mlsanity.com/paper/2604.03870/) (2026)
- [Preventing Jailbreak Prompts as Malicious Tools for Cybercriminals: A Cyber Defense Perspective](https://mlsanity.com/paper/2411.16642/) (2024)
- [Targeting Alignment: Extracting Safety Classifiers of Aligned LLMs](https://mlsanity.com/paper/2501.16534/) (2025)
- [ConceptGuard: Neuro-Symbolic Safety Guardrails via Sparse Interpretable Jailbreak Concepts](https://mlsanity.com/paper/2508.16325/) (2025)
- [Linear probes rely on textual evidence: Results from leakage mitigation studies in language models](https://mlsanity.com/paper/2509.21344/) (2025)
- [Reflections and New Directions for Human-Centered Large Language Models](https://mlsanity.com/paper/2605.06901/) (2026)
- [An Audit and Analysis of LLM-Assisted Health Misinformation Jailbreaks Against LLMs](https://mlsanity.com/paper/2508.10010/) (2025)
- [LLM in the Middle: A Systematic Review of Threats and Mitigations to Real-World LLM-based Systems](https://mlsanity.com/paper/2509.10682/) (2025)
- [MENTOR: A Metacognition-Driven Self-Evolution Framework for Uncovering and Mitigating Implicit Domain Risks in LLMs](https://mlsanity.com/paper/2511.07107/) (2025)
- [Dark LLMs: The Growing Threat of Unaligned AI Models](https://mlsanity.com/paper/2505.10066/) (2025)
- [Introducing the Generative Application Firewall (GAF)](https://mlsanity.com/paper/2601.15824/) (2026)
- [A Causal Perspective for Enhancing Jailbreak Attack and Defense](https://mlsanity.com/paper/2602.04893/) (2026)
- [Why Do Large Language Models Generate Harmful Content?](https://mlsanity.com/paper/2604.11663/) (2026)
- [Latent Adversarial Detection: Adaptive Probing of LLM Activations for Multi-Turn Attack Detection](https://mlsanity.com/paper/2604.28129/) (2026)
- [Minimal, Local, Causal Explanations for Jailbreak Success in Large Language Models](https://mlsanity.com/paper/2605.00123/) (2026)
- [Breaking Refusal in the First Half: A Mechanistic Study of the Prefill Jailbreak](https://mlsanity.com/paper/2607.14147/) (2026)
- [Measuring the Wrong Thing: Internal Harmfulness Scores Anti-Rank Successful Jailbreaks](https://mlsanity.com/paper/2608.09624/) (2026)
- [The Safety Relay in Roleplay Jailbreaks: A Component-Resolved Causal Analysis of Harm Recognition and Refusal](https://mlsanity.com/paper/2608.30585/) (2026)
- [MechAudit-40: White-Box Auditing across 40 LLM Attack Mechanisms](https://mlsanity.com/paper/2609.06612/) (2026)
- [Don't Judge an LLM Only by Its Activations: Discovering Suppressed Safety Features via Counterfactual Activation Potential](https://mlsanity.com/paper/2610.05541/) (2026)

## Abstract (arXiv)

Jailbreaks have been a central focus of research regarding the safety and reliability of large language models (LLMs), yet the mechanisms underlying these attacks remain poorly understood. While previous studies have predominantly relied on linear methods to detect jailbreak attempts and model refusals, we take a different approach by examining both linear and non-linear features in prompts that lead to successful jailbreaks. First, we introduce a novel dataset comprising 10,800 jailbreak attempts spanning 35 diverse attack methods. Leveraging this dataset, we train linear and non-linear probes on hidden states of open-weight LLMs to predict jailbreak success. Probes achieve strong in-distribution accuracy but transfer is attack-family-specific, revealing that different jailbreaks are supported by distinct internal mechanisms rather than a single universal direction. To establish causal relevance, we construct probe-guided latent interventions that systematically shift compliance in the predicted direction. Interventions derived from non-linear probes produce larger and more reliable effects than those from linear probes, indicating that features linked to jailbreak success are encoded non-linearly in prompt representations. Overall, the results surface heterogeneous, non-linear structure in jailbreak mechanisms and provide a prompt-side methodology for recovering and testing the features that drive jailbreak outcomes.