RSS

Christian Moya Calderon

Karma: 43

Ph.D. in Mathematics. Work on mechanistic interpretability and alignment.

Tie train­ing can make DPO/​RLHF-trained AIs gen­er­al­ize better

Elliott Thornley, Christian Moya Calderon and Alex Semendinger
6 Jul 2026 15:25 UTC
52 points
8 comments15 min readLW link
(arxiv.org)