Bronson Schoen

Karma: 179

Working in self-driving for about a decade. Currently at NVIDIA. Interested in opportunities to contribute to alignment.

Frontier Models are Capable of In-context Scheming

Marius Hobbhahn, AlexMeinke, Bronson Schoen, rusheb, Jérémy Scheurer and Mikita Balesni

5 Dec 2024 22:11 UTC

196 points

22 comments7 min readLW link