Lesser Wrong Feed

 Ablating 1 of a chess transformer's 128 attention heads makes the model stop finding Paul Morphy's queen sacrifice
 Thu, 20 Aug 2026 21:17:51 GMT
 Llama will abandon a correct answer if it thinks you're educated
 Thu, 20 Aug 2026 20:06:20 GMT
 Why self-fund your regranting?
 Thu, 20 Aug 2026 19:18:02 GMT
 If Aliens Exist, We Should Expect to Find Them Around Now
 Thu, 20 Aug 2026 18:57:08 GMT
 Cross-Dataset Transfer Evaluation of Deception Probes in Smaller Models
 Thu, 20 Aug 2026 17:16:51 GMT
 The Fourth Humiliation
 Thu, 20 Aug 2026 17:03:05 GMT
 AI #182: Pause For Reflection
 Thu, 20 Aug 2026 14:50:28 GMT
 We Must Remember That Our World Contains Hell
 Thu, 20 Aug 2026 14:22:26 GMT
 Thoughts on Taking OpenAI Foundation Funding
 Thu, 20 Aug 2026 13:00:34 GMT
 Making sense of the misalignment risk model in the Anthropic Risk Report (August 2026)
 Thu, 20 Aug 2026 02:33:31 GMT