collide-1m

COLLIDE-1M: COmprehensive Lhc coLlision Dataset for Foundation model dEvelopment (1 Million) See COLLIDE2V example notebook.ipynb for a quick demo, showing how to load the dataset and inspect its contents. Changelog: 09/10/25: Updated the data to fix a numerical Delphes bug and have random seeds consistent with the final set

Source: Hugging Face Hub (fastmachinelearning/collide-1m). Metadata imported from the dataset’s Hub tags.