Circuit Discovery Helps Detect LLM Jailbreaking: A Mechanistic Interpretability Study

Despite extensive safety alignment, large language models LLMs remain vulnerable to jailbreak attacks that bypass safeguards to elicit harmful content. While prior work attributes this vulnerability to safety training limitations, the internal mechanisms by which LLMs process adversarial prompts...

Amplified Vulnerabilities: Structured Jailbreak Attacks on LLM-Based Multi-Agent Debate

Multi-Agent Debate MAD, leveraging collaborative interactions among Large Language Models LLMs, aim to enhance reasoning capabilities in complex tasks. However, the security implications of their iterative dialogues and role-playing characteristics, particularly susceptibility to jailbreak attack...

Threat Source newsletter (Aug. 22)

Newsletter compiled by Jon Munshaw. Welcome to this week’s Threat Source newsletter — the perfect place to get caught up on all things Talos from the past week. What’s old is new again. Our research this week centers around a series of long-lasting threat actors and malware that have been given n...