/competition-metageneralization

Reproducible experiments to show that filtering participants improves generalization in competitions and benchmarks

Primary LanguageJupyter Notebook

Watchers