Impact of Synthetic-to-Real Sample Ratios on CausalMixFT Calibration and Generalization in Tabular Models

Fine-tuning tabular foundation models (TFMs) under data scarcity is challenging, as early stopping on even scarcer validation data often fails to capture true generalization performance. We propose CausalMixFT, a method that enhances fine-tuning robustness and downstream performance by generating structurally consistent synthetic samples using Structural Causal Models (SCMs) fitted on the target dataset. This approach augments limited real data with causally informed synthetic examples, preserving feature dependencies while expanding training diversity. Evaluated across 33 classification datas

Research goal: What is the impact of varying the ratio of synthetic-to-real samples in CausalMixFT on the calibration error and generalization performance of tabular foundation models, as measured by metrics like Brier score and AUC on public tabular benchmarks?

Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 7.7/10.