Lesser Wrong Feed

 Unsupervised Agent Discovery
 Mon, 22 Dec 2025 22:01:57 GMT
 Announcing Gemma Scope 2
 Mon, 22 Dec 2025 21:56:59 GMT
 Keeping Up Against the Joneses: Balsa’s 2025 Fundraising Letter
 Mon, 22 Dec 2025 21:32:52 GMT
 [Intro to AI Alignment] 0. Overview and Foundations
 Mon, 22 Dec 2025 21:20:45 GMT
 $500 Write like lsusr competition
 Mon, 22 Dec 2025 20:09:49 GMT
 Appendices: Supervised finetuning on low-harm reward hacking generalises to high-harm reward hacking
 Mon, 22 Dec 2025 19:33:28 GMT
 Supervised finetuning on low-harm reward hacking generalises to high-harm reward hacking
 Mon, 22 Dec 2025 19:32:12 GMT
 Recent LLMs can use filler tokens or problem repeats to improve (no-CoT) math performance
 Mon, 22 Dec 2025 17:21:09 GMT
 Can we interpret latent reasoning using current mechanistic interpretability tools?
 Mon, 22 Dec 2025 16:56:24 GMT
 Why does Eliezer make abrasive public comments?
 Mon, 22 Dec 2025 16:45:18 GMT