Home News
An auditing agent that enables automated monitoring and interaction with language models to detect potential alignment issues, reward hacking, and other concerning behaviors.
1,656 Likes
983 Followers