/ Research

Physics-Inspired Optimization Improves Deep LLM Pruning

2 days ago

OpenAI Forms Independent Group to Guide AI Mathematics

2 days ago

OpenAI’s Astra Claims Ten Major Mathematics Advances

1 month ago

Google ATLAS Maps How People Actually Use AI

2 months ago

OpenAI Proposes a Method for Evaluating Models’ Reward-Seeking Behavior

2 months ago

Anthropic Identifies New Failures in AI Agent Behavior

2 months ago

Anthropic reveals hidden signals inside Claude’s reasoning

2 months ago

Anthropic found an internal “workspace” for reasoning in Claude

2 months ago