Guide Labs debuts a new kind of interpretable LLM TechCrunch By:Tim Fernholz 23 February 2026 at 21:53 The company open sourced an 8-billion-parameter LLM, Steerling-8B, trained with a new architecture designed to make its actions easily interpretable.
OpenAI debated calling police about suspected Canadian shooterβs chats TechCrunch By:Tim Fernholz 21 February 2026 at 19:25 Jesse Van Rootselaar's descriptions of gun violence were flagged by tools that monitor ChatGPT for misuse.