FeedSourcesModels
⌘K
AR

Anthropic Research

26 articles total

Go to source

Anthropic is an AI safety and research company that's working to build reliable, interpretable, and steerable AI systems.

Go to source
  •  Yes, Claude can do Nine Loops
  •  Project Swap: What happens when agents trade for us?
  •  How Claude is uplifting biomolecular modeling
  •  Measuring tactical intelligence targeting and conventional weapons capabilities of AI models
  •  An alignment assessment of recent cybersecurity incidents
  •  Formalizing Fermat's Last Theorem
  •  Automated researchers can reliably mitigate alignment failures
  •  Enabling independent research on how people use Claude
  •  How Claude is accelerating protein design and analytical chemistry
  •  Patterns and problems in emerging multiagent systems
  •  Reviewing the evidence on worker retraining programs
  •  Learning more about Claude's mathematical capabilities
  •  Discovering cryptographic weaknesses with Claude
  •  Project Pilot: Can AI control a drone?
  •  How Canada uses Claude: Findings from the Anthropic Economic Index
  •  Claude’s values across models and languages
  •  Claude plays robotics
  •  An off switch for dual use knowledge in AI models
  •  A global workspace in language models
  •  Anthropic Economic Index report: Cadences