Back to Home

Evaluate Your Model

Submit your model for transparent, reproducible evaluation on CORE

Evaluation Process

  • • Your model will be evaluated on all 204 CORE questions
  • • We use standardized prompts to ensure fair comparison
  • • Response distributions are compared to human baselines
  • • Results are verified and reproducible
  • • Evaluation typically takes 2-3 days depending on model access

Model Submission Form

Contact Information

Model Information

By submitting, you agree to our evaluation protocols and reproducibility standards.