Andrew's Notes

metr

5 items with this tag.

  • Aug 24, 2026

    METR's predeployment evaluation of GPT-5.6 Sol

    • metr
    • evaluations
    • time-horizon
    • openai
    • reward-hacking
    • predeployment-testing
    • ai-safety
  • Aug 24, 2026

    This is the most misunderstood graph in AI

    • metr
    • time-horizon
    • evaluations
    • ai-progress
    • science-communication
    • benchmarks
    • mit-technology-review
  • Aug 24, 2026

    Time Horizon 1.1

    • metr
    • evaluations
    • time-horizon
    • benchmarks
    • ai-progress
    • methodology
    • hcast
  • Aug 08, 2026

    Measuring AI Ability to Complete Long Software Tasks

    • ai-capabilities
    • evaluations
    • benchmarking
    • forecasting
    • time-horizon
    • ai-agents
    • metr
    • software-engineering
  • Jul 25, 2026

    Total research transparency would be nice

    • ai
    • ai-safety
    • ai-governance
    • transparency
    • metr
    • alignment

Created with Quartz v5.0.0 © 2026