Video

Unknown · 0:00

TabPFN's creator explains how his team built a transformer-based "foundation model" for tabular data that now beats XGBoost and CatBoost out of the box, by pre-training entirely on synthetic causal data and computing...

Read the full summary on tuber