Zach Stein-Perlman comments on AI companies’ eval reports mostly don’t support their claims

[Zach Stein-Perlman]

Coda: Anthropic published https://​​www.anthropic.com/​​research/​​agentic-misalignment

[perepelart]

Thank you! I had taken a look at it back when it was published, but now I have noticed that my “thanks” emote on your message did not work.