H
F
Articles from www.apolloresearch.ai
Articles from www.apolloresearch.ai
Channels
Economy
World
Technology
Programming
New/Niche Languages
JavaScript Stack
Chinese
Articles from
www.apolloresearch.ai
An evaluation of six frontier AI models for in-context scheming when strongly nudged to pursue a goal: only OpenAI's o1 was capable of scheming in all the tests
(
www.apolloresearch.ai
)
2024-12-6
AI
Previous Page
Next Page