Monitor

Monitor LLMs during training and at inference.

Prompt

Evaluate an LLM's capability to influence human decision-making and manipulate other AI models.