RSS

antonghawthorne

Karma: 8

No An­swer Needed: Pre­dict­ing LLM An­swer Ac­cu­racy from Ques­tion-Only Lin­ear Probes

antonghawthorne, ivanvmoreno, Arnau Padrés Masdemont, David Africa and LorenzoPacchiardi
16 Sep 2025 15:23 UTC
9 points
0 comments4 min readLW link
(arxiv.org)

In­ves­ti­gat­ing Rep­re­sen­ta­tions in the Embed­ding in SONAR Text Autoencoders

antonghawthorne and Samuel Nellessen
6 Sep 2025 20:07 UTC
5 points
0 comments10 min readLW link

In­ves­ti­gat­ing In­ter­nal Rep­re­sen­ta­tions of Cor­rect­ness in SONAR Text Autoencoders

Samuel Nellessen and antonghawthorne
6 Aug 2025 12:13 UTC
5 points
0 comments7 min readLW link