Ken Griffin donates $3B to Carnegie Mellon as university plots Miami campusft.com·3 pts·bearseascape·0
Model Spec Midtraining: Improving How Alignment Training Generalizesalignment.anthropic.com·2 pts·bearseascape·0
Why one small American town won't stop stoning its residents to deatharchiveofourown.org·2 pts·bearseascape·1
MooseAgent: A LLM Based Multi-Agent Framework for Automating Moose Simulationarxiv.org·13 pts·bearseascape·0
Towards Understanding Distilled Reasoning Models: A Representational Approacharxiv.org·3 pts·bearseascape·0
Transformers Learn to Implement Multistep Gradient Descent with Chain of Thoughtarxiv.org·1 pts·bearseascape·0