The Geometry of Harmfulness in Multi-Turn Attacks

arXiv:2609.38389 · cs.CR, cs.CL · Submitted 2026-09-29 · Read on arXiv

cs.CR, cs.CL

Submitted: 2026-09-29

Updated: 2026-09-29

Terminology

• Cosine Similarity — 12× in this paper · explained in 3 episodes
• Representation Engineering — 9× in this paper · explained in 2 episodes
• Alignment — 7× in this paper · explained in 1 episode
• Robustness — 6× in this paper · explained in 5 episodes
• Steering — 5× in this paper · explained in 2 episodes
• Adversarial Attacks — 4× in this paper · explained in 1 episode
• Transformer — 4× in this paper · explained in 2 episodes
• Adversarial Prompting — 2× in this paper · explained in 1 episode
• Interpretability — 2× in this paper · explained in 15 episodes
• Large Language Models (LLMs) — 2× in this paper · explained in 6 episodes
• Transferability — 2× in this paper · explained in 4 episodes
• Ablation — 1× in this paper · explained in 1 episode
• Activation Steering — 1× in this paper · explained in 4 episodes
• Actor — 1× in this paper · explained in 3 episodes
• Benchmark — 1× in this paper · explained in 2 episodes
• Benchmarking — 1× in this paper · explained in 3 episodes
• Core — 1× in this paper · explained in 2 episodes
• Distribution Shift — 1× in this paper · explained in 5 episodes
• Generative Model — 1× in this paper · explained in 2 episodes
• Generative Models — 1× in this paper · explained in 2 episodes
• Latent Representation — 1× in this paper · explained in 2 episodes
• Linear probe — 1× in this paper · explained in 2 episodes

Sources

• Emerging Vulnerabilities in Frontier Models: Multi-Turn Jailbreak Attacks
• AgentLAB: Benchmarking LLM Agents against Long-Horizon Attacks
• LLM Defenses Are Not Robust to Multi-Turn Human Jailbreaks Yet
• LLMs know their vulnerabilities: Uncover Safety Gaps through Natural Distribution Shifts
• Steering Language Models With Activation Engineering
• Speak Out of Turn: Safety Vulnerability of Large Language Models in Multi-turn Dialogue
• Representation Engineering: A Top-Down Approach to AI Transparency
• Universal and Transferable Adversarial Attacks on Aligned Language Models

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs