News

inspect-petri 3.1.1

An auditing agent that enables automated monitoring and interaction with language models to detect potential alignment issues, reward hacking, and other concerning behaviors.

Continue Reading at Pypi.org →