LLM Security

Learn about vulnerabilities and defense mechanisms for LLM applications

Lesson 1

Prompt Injection

Attack vectors & defense

Learn how prompt injection attacks work with 12 real examples. Defense strategies: input validation, sandboxing and guardrails.

NewLesson 2

Jailbreaking — LLM Safety Bypass Techniques

Bypassing restrictions

Learn about techniques used to bypass LLM safety measures and how to prevent them

NewLesson 3

Factuality & Hallucinations

When LLMs make things up

Understand why LLMs hallucinate and strategies to improve factual accuracy

NewLesson 4

Biases in LLMs

Unfair outputs

Explore biases in language models and methods to detect and mitigate them