1 articles with this tag
BeyondArena tests 11 models on 142 datasets and finds tabular foundation models lead on small IID data but trail GBDTs on large and non-IID tasks.
