Request for Releasing Google Research Football (GRF) Experiment Code

[lexvvcy]

Dear authors, thanks for the great work on M3HF!

We noticed that the GRF 5v5 HARD experiments (Appendix Figures 5–6) were added in the rebuttal session and have not yet included in the current repo. Do you plan to release the GRF training/evaluation code or configs used for those results?

If possible, we would really appreciate an early or partial release, as we are hoping to use these experiments soon. This would be very helpful for reproducibility and timely comparisons. Thanks a lot!

[ziyan-wang98]

Thanks for your interest in our work. At the moment, we don’t plan to release the GRF-related code due to code privacy constraints; however, you can still train and reproduce the core setup with M3HF using the public DB-Football repository: https://github.com/Shanghai-Digital-Brain-Laboratory/DB-Football.