RSS

Monte M

Karma: 1,100

Sleeper Agents: Train­ing De­cep­tive LLMs that Per­sist Through Safety Training

12 Jan 2024 19:51 UTC
288 points
94 comments3 min readLW link
(arxiv.org)

Paper: Un­der­stand­ing and Con­trol­ling a Maze-Solv­ing Policy Network

13 Oct 2023 1:38 UTC
69 points
0 comments1 min readLW link
(arxiv.org)

Ac­tAdd: Steer­ing Lan­guage Models with­out Optimization

6 Sep 2023 17:21 UTC
105 points
3 comments2 min readLW link
(arxiv.org)

Open prob­lems in ac­ti­va­tion engineering

24 Jul 2023 19:46 UTC
43 points
2 comments1 min readLW link
(coda.io)

Steer­ing GPT-2-XL by adding an ac­ti­va­tion vector

13 May 2023 18:42 UTC
416 points
97 comments50 min readLW link

Un­der­stand­ing and con­trol­ling a maze-solv­ing policy network

11 Mar 2023 18:59 UTC
312 points
22 comments23 min readLW link