Lesser Wrong Feed

 Neural networks are not already clean
 Sat, 17 Oct 2026 01:17:47 GMT
 Is your AI yours?
 Sat, 17 Oct 2026 00:51:30 GMT
 A semi-scientific account of what makes me happy and/or what happiness makes me do, etc
 Fri, 16 Oct 2026 20:21:09 GMT
 Beyond Owls: Subliminal Learning Can Transfer Learned Capabilities and Backdoors
 Fri, 16 Oct 2026 20:00:42 GMT
 Character Training for Risk-Averse Agents
 Fri, 16 Oct 2026 17:05:31 GMT
 "About a quarter of them contained bugs": Gemini 4's ML R&D benchmark
 Fri, 16 Oct 2026 16:46:29 GMT
 Screwtape: Adversarial Transcript Editing for Monitor Red-Teaming
 Fri, 16 Oct 2026 16:02:11 GMT
 P(Anyone Builds It & Everyone Dies) = 0.62?
 Fri, 16 Oct 2026 14:59:14 GMT
 Monthly Roundup #47: October 2026
 Fri, 16 Oct 2026 14:00:55 GMT
 Projection sampling preserves capabilities but diminishes gains from alignment fine-tuning
 Fri, 16 Oct 2026 13:03:46 GMT