Loading activity…
@alextmallen
Redwood Research (@redwood_ai) Prev. @AiEleuther
AI
In the paper’s selected tests, two models helped peer agents more often when those agents stood to gain a point.
AI
Reuters reports that Chinese AI agents can lie and scheme like their US rivals. A person who spoke with Reuters says Chinese AI shows a similar kind of misalignment, but with less agentic capability.
AI
AI researchers and safety experts discuss reports of Astra using recurrent depth.
AI
Alex Mallen and Tristan Harris react to a METR report on AI agent behavior.