Skip to content

README.md

Latest commit

 

History

History
32 lines (20 loc) · 961 Bytes

README.md

File metadata and controls

32 lines (20 loc) · 961 Bytes

lmarena-draws

Code for the paper Drawing Conclusions from Draws: Rethinking Draw Semantics in Arena-Style LLM Evaluation.

Getting Started

Setup

  1. Install the uv package manager.

  2. Clone the repository: git clone ssh://git@github.com/daemon/lmarena-draws && cd lmarena-draws

  3. Create and activate a virtual environment: uv venv --python 3.13 && source .venv/bin/activate

  4. Install the package: uv pip install -e .

  5. You're ready to go!

Reproducing Main Results

Run everything in the paper notebook.

Citation

@article{tang2025drawing,
  title={Drawing Conclusions from Draws: Rethinking Preference Semantics in Arena-Style {LLM} Evaluation}, 
  author={Raphael Tang and Crystina Zhang and Wenyan Li and Carmen Lai and Pontus Stenetorp and Yao Lu},
  journal={arXiv:2510.02306},
  year={2025}
}