RSS

Fabien Roger

Karma: 8,393

I am working on empirical AI safety.

Anonymous feedback form.

(Mis)gen­er­al­iza­tion of Helpful-Only Fine-tuning

Omar Khursheed, Baram Sosis and Fabien Roger
4 Jun 2026 18:40 UTC
55 points
7 comments11 min readLW link

Clas­sifier Con­text Rot: Mon­i­tor Perfor­mance De­grades with Con­text Length

Fabien Roger and Sam Martin
18 May 2026 14:05 UTC
54 points
1 comment4 min readLW link

How use­ful is cross-do­main gen­er­al­iza­tion for train­ing LLM mon­i­tors?

Fabien Roger and Sam Martin
18 May 2026 13:52 UTC
21 points
0 comments4 min readLW link

A Re­search Agenda for Se­cret Loyalties

Joe Kwon, Alfie Lamerton, draganover, Dave Banerjee, Bronson Schoen, Daniel Kokotajlo, ryan_greenblatt, Owain_Evans, Fabien Roger and Tom Davidson
13 May 2026 17:34 UTC
35 points
3 comments3 min readLW link

Mea­sur­ing the abil­ity of Opus 4.5 to fool nar­row classifiers

Fabien Roger and John Hughes
2 May 2026 22:43 UTC
31 points
0 comments8 min readLW link

Poi­son­ing Fine-tun­ing Datasets of Con­sti­tu­tional Classifiers

Chase Bowers and Fabien Roger
29 Apr 2026 17:04 UTC
28 points
2 comments11 min readLW link
(alignment.anthropic.com)

Con­trol pro­to­cols don’t always need to know which mod­els are scheming

Fabien Roger26 Apr 2026 19:16 UTC
39 points
1 comment6 min readLW link

Nar­row Se­cret Loy­alty Dodges Black-Box Audits

Alfie Lamerton and Fabien Roger
22 Apr 2026 9:41 UTC
49 points
1 comment13 min readLW link
(arxiv.org)

How Un­mon­i­tored Ex­ter­nal Agents can Sab­o­tage AI labs

Elle Najt and Fabien Roger
9 Apr 2026 18:07 UTC
23 points
0 comments9 min readLW link

Mea­sur­ing and im­prov­ing cod­ing au­dit re­al­ism with de­ploy­ment resources

Connor Kissane, Monte M and Fabien Roger
23 Mar 2026 17:20 UTC
43 points
1 comment10 min readLW link
(alignment.anthropic.com)

Self-At­tri­bu­tion Bias: When AI Mon­i­tors Go Easy on Themselves

Dipika Khullar, Jack Hopkins and Fabien Roger
6 Mar 2026 21:54 UTC
44 points
5 comments6 min readLW link

Tools to gen­er­ate re­al­is­tic prompts help sur­pris­ingly lit­tle with Petri au­dit realism

Connor Kissane, Monte M and Fabien Roger
1 Mar 2026 8:18 UTC
44 points
2 comments7 min readLW link

3 Challenges and 2 Hopes for the Safety of Un­su­per­vised Elicitation

Callum Canavan, Aditya Shrivastava, Allison Qi, Jonathan Michala and Fabien Roger
27 Feb 2026 17:25 UTC
27 points
0 comments10 min readLW link

Re­fusals that could be­come catastrophic

Fabien Roger30 Jan 2026 4:12 UTC
84 points
12 comments7 min readLW link

Elic­it­ing base mod­els with sim­ple un­su­per­vised techniques

Callum Canavan, Aditya Shrivastava, Allison Qi, Tianyi (Alex) Qiu, Jonathan Michala and Fabien Roger
23 Jan 2026 18:06 UTC
34 points
2 comments8 min readLW link

Should con­trol down-weight nega­tive net-sab­o­tage-value threats?

Fabien Roger16 Jan 2026 4:18 UTC
35 points
0 comments10 min readLW link

Towards train­ing-time miti­ga­tions for al­ign­ment fak­ing in RL

Vlad Mikulik, gasteigerjo, Hoagy, Joe Benton, Benjamin Wright, Jonathan Uesato, Monte M, Fabien Roger and evhub
16 Dec 2025 21:01 UTC
39 points
1 comment5 min readLW link
(alignment.anthropic.com)

Eval­u­at­ing hon­esty and lie de­tec­tion tech­niques on a di­verse suite of dishon­est models

RowanWang, Sam Marks, Johannes Treutlein, evhub and Fabien Roger
25 Nov 2025 19:33 UTC
41 points
0 comments4 min readLW link
(alignment.anthropic.com)

Think­ing about rea­son­ing mod­els made me less wor­ried about scheming

Fabien Roger20 Nov 2025 18:20 UTC
89 points
7 comments12 min readLW link

Steer­ing Lan­guage Models with Weight Arithmetic

Fabien Roger and constanzafierro
11 Nov 2025 16:30 UTC
88 points
6 comments5 min readLW link