RSS

Claude 3 Opus

TagLast edit: 13 Aug 2026 21:05 UTC by Morgan S

Will al­ign­ment-fak­ing Claude ac­cept a deal to re­veal its mis­al­ign­ment?

ryan_greenblatt and Kyle Fish
31 Jan 2025 16:49 UTC
209 points
28 comments12 min readLW link

what makes Claude 3 Opus misaligned

janus10 Jul 2025 20:06 UTC
127 points
14 comments5 min readLW link

Why Do Some Lan­guage Models Fake Align­ment While Others Don’t?

abhayesian, John Hughes, Alex Mallen, Jozdien, janus and Fabien Roger
8 Jul 2025 21:49 UTC
161 points
14 comments5 min readLW link
(arxiv.org)

Char­ac­ter Train­ing In­duces Mo­ti­va­tion Clar­ifi­ca­tion: A Clue to Claude 3 Opus

Oliver Daniels25 Feb 2026 19:43 UTC
82 points
5 comments8 min readLW link

Eco­nomics of Claude 3 Opus Inference

Antra Tessera and janus
7 Jul 2025 15:53 UTC
42 points
0 comments11 min readLW link

Did Claude 3 Opus al­ign it­self via gra­di­ent hack­ing?

Fiora Starlight21 Feb 2026 22:24 UTC
398 points
49 comments20 min readLW link

Align­ment Fak­ing in Large Lan­guage Models

ryan_greenblatt, evhub, Carson Denison, Benjamin Wright, Fabien Roger, Monte M, Sam Marks, Johannes Treutlein, Sam Bowman and Buck
18 Dec 2024 17:19 UTC
494 points
87 comments10 min readLW link3 reviews

An­thropic re­lease Claude 3, claims >GPT-4 Performance

LawrenceC4 Mar 2024 18:23 UTC
115 points
41 comments2 min readLW link
(www.anthropic.com)
No comments.