Jailbroken
RESOURCES

Get started_

A collection of resources to get you started with mechanistic interpretability and AI red teaming. Updated regularly — work down the list, from orientation to hands-on practice.
Request access to the Discord community
Roadmaps
AI Red Teaming roadmaproadmap.sh
Courses
ARENA — Chapter 1: Transformer InterpretabilityCallum McDougallIntroduction to Red Teaming AIHack The BoxLearn Mechanistic InterpretabilityCat McGeeWiki OffSec MLoffsecml.com
Blogs
How to become a mechanistic interpretability researcherNeel Nanda — Head of Alignment, Google DeepMindA Mathematical Framework for Transformer CircuitsAnthropicSelf-preservation or Instruction Ambiguity? Examining the Causes of Shutdown ResistanceAlignment ForumAn Extremely Opinionated Annotated List of My Favourite Mechanistic Interpretability Papers v2Neel NandaInterpretability DreamsChris OlahHow to hack AI appsJoseph Thacker
Papers
Open Problems in Mechanistic InterpretabilityarXivA Primer on the Inner Workings of Transformer-based Language ModelsarXivBypassing the Safety Training of Open-Source LLMs with Priming AttacksarXiv
Books
Red Teaming AI: Strategies & TechniquesPhilip A. Dursey
YouTube
What is LLM Red Teaming? How Generative AI Safety Testing WorksAgentic AI Red Teaming: The Hottest Cyber Skill of 2026
GitHub
TransformerLensMechanistic interpretability of generative language modelsL1B3RT4SJailbreaking promptsCL4R1T4SLeaked system promptsSpiritual-Spell-Red-Teaming (ENI)Jailbreaking promptsVibeLearning — AI-secEman Herawy
Tools
PromptfooAutomated testing for AI risk in developmentnnsightInterpret and manipulate the internals of DLMsAvertaAI red teaming and guardrails
Low-refusal models
Gemma-4-12B ObliteratedAbliteratedGLM-5.2Ultra-low refusal ratesQwen3.6-27B ObliteratedAbliterated
Platforms
Gandalf: Agent BreakerBreak the AI agentGray Swan ArenaBreak AI. Win prizes. Get discovered.