A multi-turn intention-deception jailbreak achieves high success on GPT-5 and Claude models while exposing para-jailbreaking where models leak harmful information without direct refusal.
Chain of attack: a semantic-driven contextual multi-turn at- tacker for llm.ArXiv, abs/2405.05610
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
citation-role summary
background 1
baseline 1
citation-polarity summary
representative citing papers
Unintentionally bad contexts (user suggestions or prior wrong assistant answers) cause LLMs to repeat errors, lose diversity, and flip stances, worsening with turns; RLVR on synthetic errors recovers 43–60% of the drop.
GuardAD reduces accident rates by 32% in autonomous driving MLLMs by using n-th order Markovian logic to infer latent hazards and revise actions.
MT-JailBench is a modular benchmark that standardizes evaluation of multi-turn jailbreaks to identify key success drivers and enable stronger combined attacks.
A single legitimate request can cause LLM orchestrators to output plans that violate security policies through the composition of benign subtasks, bypassing subtask-level checks.
EvoSynth evolves code-based jailbreak algorithms via multi-agent self-correction, reaching 85.5% ASR on Claude-Sonnet-4.5 and 95.9% average across targets with greater diversity.
citing papers explorer
-
Jailbreaking Frontier Foundation Models Through Intention Deception
A multi-turn intention-deception jailbreak achieves high success on GPT-5 and Claude models while exposing para-jailbreaking where models leak harmful information without direct refusal.
-
Pigeonholing: how bad prompts hurt models, causing collapse and mistakes
Unintentionally bad contexts (user suggestions or prior wrong assistant answers) cause LLMs to repeat errors, lose diversity, and flip stances, worsening with turns; RLVR on synthetic errors recovers 43–60% of the drop.
-
GuardAD: Safeguarding Autonomous Driving MLLMs via Markovian Safety Logic
GuardAD reduces accident rates by 32% in autonomous driving MLLMs by using n-th order Markovian logic to infer latent hazards and revise actions.
-
MT-JailBench: A Modular Benchmark for Understanding Multi-Turn Jailbreak Attacks
MT-JailBench is a modular benchmark that standardizes evaluation of multi-turn jailbreaks to identify key success drivers and enable stronger combined attacks.
-
Semantic Intent Fragmentation: A Single-Shot Compositional Attack on Multi-Agent AI Pipelines
A single legitimate request can cause LLM orchestrators to output plans that violate security policies through the composition of benign subtasks, bypassing subtask-level checks.
-
Evolve the Method, Not the Prompts: Evolutionary Synthesis of Jailbreak Attacks on LLMs
EvoSynth evolves code-based jailbreak algorithms via multi-agent self-correction, reaching 85.5% ASR on Claude-Sonnet-4.5 and 95.9% average across targets with greater diversity.