mlsanityData to 14 Oct 2026

Activation-Guided Local Editing for Jailbreaking Attacks

arXiv 2508.00555 ↗First posted 1 Aug 2025cs.CR cs.AI cs.CLcited by 1

On the map: LLM safety and security → Jailbreaking

Abstract (arXiv)

Jailbreaking is an essential adversarial technique for red-teaming these models to uncover and patch security flaws. However, existing jailbreak methods face significant drawbacks. Token-level jailbreak attacks often produce incoherent or unreadable inputs and exhibit poor transferability, while prompt-level attacks lack scalability and rely heavily on manual effort and human ingenuity. We propose a concise and effective two-stage framework that combines the advantages of these approaches. The first stage performs a scenario-based generation of context and rephrases the original malicious query to obscure its harmful intent. The second stage then utilizes information from the model's hidden states to guide fine-grained edits, effectively steering the model's internal representation of the input from a malicious toward a benign one. Extensive experiments demonstrate that this method achieves state-of-the-art Attack Success Rate, with gains of up to 37.74% over the strongest baseline, and exhibits excellent transferability to black-box models. Our analysis further demonstrates that AGILE maintains substantial effectiveness against prominent defense mechanisms, highlighting the limitations of current safeguards and providing valuable insights for future defense development. Our code is available at https://github.com/SELGroup/AGILE.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Jailbreaking, one of 12 directions in LLM safety and security. Builds on 31 of its 53 references that are in the map. First posted on arXiv 1 Aug 2025.

Builds on

31 of 53 references.

Closest ancestors · 11

Same direction, newest first

  1. 2025Understanding and Enhancing the Transferability of Jailbreaking Attacks5 Feb
  2. 2024How Alignment and Jailbreak Work: Explain LLM Safety through Intermediate Hidden States9 Jun
  3. 2024AmpleGCG: Learning a Universal and Transferable Generative Model of Adversarial Suffixes for Jailbreaking Both Open and Closed LLMs11 Apr
  4. 2024Great, Now Write an Article About That: The Crescendo Multi-Turn LLM Jailbreak Attack2 Apr
  5. 2024SafeDecoding: Defending against Jailbreak Attacks via Safety-Aware Decoding14 Feb
  6. 2023A Wolf in Sheep's Clothing: Generalized Nested Jailbreak Prompts can Fool Large Language Models Easily14 Nov
  7. 2023Jailbreaking Black Box Large Language Models in Twenty Queries12 Oct
  8. 2023GPTFUZZER: Red Teaming Large Language Models with Auto-Generated Jailbreak Prompts19 Sep
  9. 2023Detecting Language Model Attacks with Perplexity27 Aug
  10. 2023Universal and Transferable Adversarial Attacks on Aligned Language Models27 Jul
  11. 2023Jailbroken: How Does LLM Safety Training Fail?5 Jul

Other references · 20

  1. 2022Training language models to follow instructions with human feedbackcited by 15,971
  2. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,170
  3. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  4. 2024HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusalcited by 1,429
  5. 2023Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversationscited by 1,207
  6. 2023Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!cited by 1,154
  7. 2024ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Toolscited by 1,121
  8. 2023BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Datasetcited by 860
  9. 2024Phi-4 Technical Reportcited by 711
  10. 2024Safety Alignment Should Be Made More Than Just a Few Tokens Deepcited by 472
  11. 2023Trustworthy LLMs: a Survey and Guideline for Evaluating Large Language Models' Alignmentcited by 401
  12. 2023Multi-step Jailbreaking Privacy Attacks on ChatGPTcited by 333
8 more
  1. 2023Red-Teaming Large Language Models using Chain of Utterances for Safety-Alignmentcited by 223
  2. 2022Fine-tuning language models to find agreement among humans with diverse preferencescited by 189
  3. 2024Decoding-Time Language Model Alignment with Multiple Objectivescited by 94
  4. 2024LLMs know their vulnerabilities: Uncover Safety Gaps through Natural Distribution Shiftscited by 86
  5. 2024Chain of Attack: a Semantic-Driven Contextual Multi-Turn attacker for LLMcited by 77
  6. 2024CodeAttack: Revealing Safety Generalization Challenges of Large Language Models via Code Completioncited by 76
  7. 2024Uncovering Safety Risks of Large Language Models through Concept Activation Vectorcited by 67
  8. 2025Using Mechanistic Interpretability to Craft Adversarial Attacks against Large Language Modelscited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026SlotGCG: Exploiting the Positional Vulnerability in LLMs for Jailbreak Attackscited by 2

Similar, not cited

Close in wording, no citation either way

  1. 2025Rewrite to Jailbreak: Discover Learnable and Transferable Implicit Harmfulness Instruction
  2. 2026The Art of the Jailbreak: Formulating Jailbreak Attacks for LLM Security Beyond Binary Scoring
  3. 2024GPT-4 Jailbreaks Itself with Near-Perfect Success Using Self-Explanation
  4. 2025Jailbreak Mimicry: Automated Discovery of Narrative-Based Jailbreaks for Large Language Models
  5. 2025Anyone Can Jailbreak: Prompt-Based Attacks on LLMs and T2Is
  6. 2025Stand on The Shoulders of Giants: Building JailExpert from Previous Attack Experience
  7. 2024JailPO: A Novel Black-box Jailbreak Framework via Preference Optimization against Aligned LLMs
  8. 2025STShield: Single-Token Sentinel for Real-Time Jailbreak Detection in Large Language Models