TabPFGen – Tabular Data Generation with TabPFN

Junwei Ma, Apoorv Dankar, George Stein, Guangwei Yu, Anthony Caterini

Keywords: tabular data, in-context learning, energy based model, generative model, tabpfn

TL;DR: TabPFGen creates an energy-based generative model for tabular data using the discriminative, in-context learning capabilities of TabPFN

Abstract: Advances in deep generative modelling have not translated well to tabular data. We argue that this is caused by a mismatch in structure between popular generative models and _discriminative_ models of tabular data. We thus devise a technique to turn TabPFN -- a highly performant transformer initially designed for in-context discriminative tabular tasks -- into an energy-based generative model, which we dub _TabPFGen_. This novel framework leverages the pre-trained TabPFN as part of the energy function and does not require any additional training or hyperparameter tuning, thus inheriting TabPFN's in-context learning capability. We can sample from TabPFGen analogously to other energy-based models. We demonstrate strong results on standard generative modelling tasks, including data augmentation, class-balancing, and imputation, unlocking a new frontier of tabular data generation.

Submission Number: 3