AI is powerful when used intelligently.
Andy Agent Lab investigates the mechanisms behind AI’s human-like behavior—and applies that understanding to how intelligent systems are built, evaluated, used, and secured.
Selected writing
All articles-
Model training and evaluation
Theory of Mind in LLMs
Read article -
Model training and evaluation
Can AI Lie?
Read article -
Agent security
Role-Play Jailbreaks
Read article
Knowledge hub
Andy Agent Lab’s public knowledge hub connects research on AI behavior and security with practical guides, technical resources, and reusable prompts.
Explore the knowledge hub