Track F: Security & Adversarial ML

Red-teaming, robustness, attack/defense.

Reading Order

1. * MITRE ATLAS resource
2. Intriguing Properties of Neural Networks paper
3. Explaining and Harnessing Adversarial Examples paper
4. Universal Adversarial Triggers paper

Branches

After universal-adversarial-triggers, explore:

• Ignore Previous Prompt paper
• Not What You've Signed Up For paper