Tabular foundation models improve value prediction with longer training
A JEPA Recipe for Tabular Foundation Models
Machine LearningArtificial Intelligence
Summary
Predicting missing values in tables is important for many tasks. The authors studied two ways to train models that predict these values: one that learns by predicting actual table values and another more complex approach called JEPA. They found that JEPA needs special training steps to avoid collapsing and requires more time to converge. In large tests over many datasets, JEPA performed slightly worse than the simpler value-only method, especially for regression tasks.
What this means in practice
- •For data scientists: Improve training of tabular models by applying a JEPA recipe that stabilizes latent features and controls training length.
- •For machine learning engineers: Optimize predictive modeling pipelines for tabular data by choosing training schemes that balance accuracy and convergence time.
Authors
Mingyu Jeon, Suwan Cho, Jae Young Suh
Abstract
Tabular foundation models learn to predict cell values in context, whereas world-model self-supervision asks for prediction in representation space (LeCun, 2022; Assran et al., 2023). On a tabular foundation-model prior, the latent term of a joint-embedding predictive architecture (JEPA) collapsed in our earlier runs and took the encoder with it to a constant map. We report a recipe under which the latent term survives to convergence beside the value objective: the value head reads the encoder field rather than the predictor, and the target is an exponential moving average (EMA) difference. To bound its cost against the value-only arm, both arms train until a plateau rule stops them, with no fixed step budget. A fixed horizon had confounded a slowdown with a ceiling, since the value-only arm was still improving well past the usual budget. At convergence, in one run per arm, the JEPA arm trails the value-only arm across 147 real datasets, 32:70 wins to losses on classification (29:63 with one entry per dataset name) and 8:24 on regression, the margin small on classification and wider on regression, and the count leans the same way in each stratum and each benchmark. The JEPA arm (jepa) needs 1.42 times as many steps as the value-only arm (ds), and 1.66 times its wall-clock, to reach its plateau.