RSS

Alexandre Variengien

Karma: 641

Bird’s eye view: An in­ter­ac­tive rep­re­sen­ta­tion to see large col­lec­tion of text “from above”.

Alexandre VariengienDec 21, 2024, 12:15 AM
10 points
4 comments5 min readLW link
(alexandrevariengien.com)

My guess at Con­jec­ture’s vi­sion: trig­ger­ing a nar­ra­tive bifurcation

Alexandre VariengienFeb 6, 2024, 7:10 PM
75 points
12 comments16 min readLW link

The case for train­ing fron­tier AIs on Sume­rian-only corpus

Jan 15, 2024, 4:40 PM
130 points
16 comments3 min readLW link

A Univer­sal Emer­gent De­com­po­si­tion of Retrieval Tasks in Lan­guage Models

Dec 19, 2023, 11:52 AM
84 points
3 comments10 min readLW link
(arxiv.org)

Cap­ture the Flag Mechanis­tic In­ter­pretabil­ity Challenges

Sep 8, 2023, 11:00 PM
24 points
0 comments7 min readLW link

In­put Swap Graphs: Dis­cov­er­ing the role of neu­ral net­work com­po­nents at scale

Alexandre VariengienMay 12, 2023, 9:41 AM
92 points
0 comments33 min readLW link

An in­tro­duc­tion to lan­guage model interpretability

Alexandre VariengienApr 20, 2023, 10:22 PM
14 points
0 comments9 min readLW link

Some com­mon con­fu­sion about in­duc­tion heads

Alexandre VariengienMar 28, 2023, 9:51 PM
64 points
4 comments5 min readLW link

Gliders in Lan­guage Models

Alexandre VariengienNov 25, 2022, 12:38 AM
30 points
11 comments10 min readLW link

Some Les­sons Learned from Study­ing Indi­rect Ob­ject Iden­ti­fi­ca­tion in GPT-2 small

Oct 28, 2022, 11:55 PM
101 points
9 comments9 min readLW link2 reviews
(arxiv.org)

Ap­ply to the Ma­chine Learn­ing For Good boot­camp in France

Alexandre VariengienJun 17, 2022, 7:32 AM
10 points
0 comments1 min readLW link

Croe­sus, Cer­berus, and the mag­pies: a gen­tle in­tro­duc­tion to Elic­it­ing La­tent Knowledge

Alexandre VariengienMay 27, 2022, 5:58 PM
17 points
0 comments16 min readLW link