Explore the code and pretrained weights for Causilo, our foundation model for classification and regression.
A foundation model for tables
Causilo is Nums AI’s tabular foundation model for classification and regression. Pretrained across many datasets, it uses examples with known outcomes as context to predict an outcome for a new row.
The same pretrained model can work with different tables without retraining for each task. Causilo provides a familiar scikit-learn interface for exploring these predictions within existing machine learning workflows.
Explore Causilo
The code and pretrained model weights are available on GitHub and Hugging Face. Our Use Cases page brings together TabArena datasets across eight domains, pairing descriptions of the prediction tasks with model comparisons.