You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

DiscoverPhysics is a benchmark for evaluating LLM agents on open-ended scientific discovery in simulated worlds with non-canonical physics. Access to the full benchmark suite, including the 11 private worlds and their evaluation rubrics, is gated to preserve the validity of the benchmark. By requesting access, you agree to the following terms:

  1. You will not publish, redistribute, or post the private world
    definitions, ground-truth force laws, or evaluation rubrics in any
    public venue (including GitHub, arXiv, blog posts, or social media).
  2. You will not use the private worlds or rubrics as training or
    fine-tuning data for any language model.
  3. You will cite the DiscoverPhysics paper in any work that uses this
    benchmark.
  4. You will report benchmark results honestly and reproducibly.

Access requests are reviewed manually and typically processed within a few business days.

Log in or Sign Up to review the conditions and access this model content.

Gated model
You can list files but not access them

Preview of files found in this repository