Lesser Wrong Feed

 Neural chameleons can('t) hide from activation oracles
 Thu, 22 Jan 2026 01:47:41 GMT
 Dedicated continuous supervision of AI companies
 Thu, 22 Jan 2026 01:52:57 GMT
 Uncovering Unfaithful CoT in Deceptive Models
 Thu, 22 Jan 2026 01:51:47 GMT
 Claude's Constitution is an excellent guide for humans, too
 Thu, 22 Jan 2026 01:26:15 GMT
 The first type of transformative AI?
 Wed, 21 Jan 2026 23:47:54 GMT
 How (and why) to read Drexler on AI
 Wed, 21 Jan 2026 23:25:59 GMT
 Finding Yourself in Others
 Wed, 21 Jan 2026 23:22:33 GMT
 AI Risks Slip Out of Mind
 Wed, 21 Jan 2026 22:54:32 GMT
 When should we train against a scheming monitor?
 Wed, 21 Jan 2026 20:48:17 GMT
 Claude Codes #3
 Wed, 21 Jan 2026 19:50:51 GMT