RSS

Claude and Perfor­ma­tive Uncertainty

Stephen Martin22 Aug 2026 19:43 UTC
15 points
0 comments12 min readLW link

When would you leave An­thropic? Notes from a chat with a ca­pa­bil­ities researcher

TheManxLoiner22 Aug 2026 18:00 UTC
24 points
0 comments3 min readLW link
(lovkush.substack.com)

Study 2 Regis­tra­tion: Ex­plor­ing rep­re­sen­ta­tional coun­ter­parts of welfare-rele­vant in­di­ca­tors un­der post-train­ing quantization

ashesfall22 Aug 2026 18:00 UTC
8 points
0 comments17 min readLW link

“Farm strength” vs “breath aware­ness”

jimmy22 Aug 2026 17:30 UTC
33 points
2 comments4 min readLW link
(beneathpsychology.com)

5 Things I Learned About Peo­ple From Do­ing Stand-Up Comedy

Luise Woehlke22 Aug 2026 15:45 UTC
15 points
1 comment6 min readLW link
(open.substack.com)

The In­stru­men­tal Con­ver­gence of Crowds

julius vidal22 Aug 2026 13:20 UTC
11 points
1 comment2 min readLW link

Hu­mans Are Align­ment Generators

James Stephen Brown22 Aug 2026 4:07 UTC
3 points
2 comments4 min readLW link
(nonzerosum.games)

Selec­tion for Selectabil­ity: In­duc­tive Bi­ases in Evolu­tion and in Neu­ral Networks

CarolusRenniusVitellius22 Aug 2026 4:02 UTC
39 points
2 comments11 min readLW link
(charlesr-w.github.io)

Align­ment fine-tun­ing in­duces con­di­tional mis­al­ign­ment in Qwen2.5-7B-In­struct

Rhea Srivats21 Aug 2026 20:40 UTC
10 points
1 comment13 min readLW link

When is Un­limited Op­ti­miza­tion Catas­trophic?

Winter Cross21 Aug 2026 20:08 UTC
33 points
0 comments13 min readLW link
(arxiv.org)

Eval­u­at­ing Ex­pla­na­tions of LLM Be­hav­ior In The Wild with Coun­ter­fac­tual Experiments

21 Aug 2026 19:09 UTC
56 points
2 comments7 min readLW link

In Defense of ASI Socialism

cguth721 Aug 2026 18:40 UTC
7 points
3 comments2 min readLW link

Misal­igned AI in the Bronze Age

frmsaul21 Aug 2026 16:51 UTC
54 points
6 comments7 min readLW link

The im­posters among us: func­tion vec­tors that ace ev­ery check and do the wrong task (in search of cir­cu­lar­ity)

star2vec21 Aug 2026 15:54 UTC
7 points
0 comments10 min readLW link

When Models Iden­tify as a Swarm

julius vidal21 Aug 2026 11:49 UTC
57 points
4 comments5 min readLW link

My Neel Nanda MATS 10.0 Ap­pli­ca­tion: Study­ing Fea­ture Split­ting in SAEs via Train­ing Data Attribution

J Rosser21 Aug 2026 9:41 UTC
17 points
0 comments13 min readLW link

Creativity Beyond the Manifold

Kartikay Luthra20 Aug 2026 23:31 UTC
8 points
2 comments6 min readLW link

Llama will aban­don a cor­rect an­swer if it thinks you’re educated

Nick Merrill20 Aug 2026 20:06 UTC
27 points
6 comments2 min readLW link

If Aliens Ex­ist, We Should Ex­pect to Find Them Around Now

jehan20 Aug 2026 18:57 UTC
12 points
1 comment1 min readLW link
(www.jehanazad.com)

We Must Re­mem­ber That Our World Con­tains Hell

James Brobin20 Aug 2026 14:22 UTC
124 points
39 comments3 min readLW link