AI Dynamics

Global AI News Aggregator

About

Frontier Model Auditing With Red-Teaming and Evaluation Agents

Our agents are useful for frontier model auditing: 1. Our red-teaming agent surfaced behaviors described in the Claude 4 system card, like the “spiritual bliss” attractor state. https://
anthropic.com/claude-4-syste
m-card
… 2. Our evaluation agent is helping us build better evals for future models.

→ View original post on X — @anthropicai,