Mitra: Mixed Synthetic Priors for Enhancing Tabular Foundation Models

• URL: http://arxiv.org/abs/2510.21204v1
• Date: Fri, 24 Oct 2025 07:15:06 GMT
• Title: Mitra: Mixed Synthetic Priors for Enhancing Tabular Foundation Models
• Authors: Xiyuan Zhang, Danielle C. Maddix, Junming Yin, Nick Erickson, Abdul Fatir Ansari, Boran Han, Shuai Zhang, Leman Akoglu, Christos Faloutsos, Michael W. Mahoney, Cuixiong Hu, Huzefa Rangwala, George Karypis, Bernie Wang,
• Abstract summary: We introduce Mitra, a TFM trained on a curated mixture of synthetic priors selected for their diversity, distinctiveness, and performance on real-world data.<n>Mitra consistently outperforms state-of-the-art TFMs, such as TabPFNv2 and TabICL, across both classification and regression benchmarks.
• Score: 85.64873567417396
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Since the seminal work of TabPFN, research on tabular foundation models (TFMs) based on in-context learning (ICL) has challenged long-standing paradigms in machine learning. Without seeing any real-world data, models pretrained on purely synthetic datasets generalize remarkably well across diverse datasets, often using only a moderate number of in-context examples. This shifts the focus in tabular machine learning from model architecture design to the design of synthetic datasets, or, more precisely, to the prior distributions that generate them. Yet the guiding principles for prior design remain poorly understood. This work marks the first attempt to address the gap. We systematically investigate and identify key properties of synthetic priors that allow pretrained TFMs to generalize well. Based on these insights, we introduce Mitra, a TFM trained on a curated mixture of synthetic priors selected for their diversity, distinctiveness, and performance on real-world tabular data. Mitra consistently outperforms state-of-the-art TFMs, such as TabPFNv2 and TabICL, across both classification and regression benchmarks, with better sample efficiency.

Related papers

• PluRel: Synthetic Data unlocks Scaling Laws for Relational Foundation Models [51.42043158297229]
  We introduce Pluel, a framework to synthesize multi-tabular relational databases from scratch.<n>In a step-by-step fashion, Pluel models (1) schemas with directed graphs, (2) inter-table primary-foreign key connectivity with bipartite graphs, and, (3) feature distributions in tables via conditional causal mechanisms.
  arXiv  Detail & Related papers  (2026-02-03T21:35:18Z)
• Robust Tabular Foundation Models [0.7539295827164078]
  A key finding is that TFMs can be pretrained entirely on synthetic datasets.<n>We introduce an optimality gap measure, given by the difference between TFM performance and the best achievable performance.<n>These results highlight a promising new dataset for targeted adversarial training and fine-tuning of TFMs using synthetic data alone.
  arXiv  Detail & Related papers  (2025-12-02T23:40:39Z)
• Scaling Laws of Synthetic Data for Language Models [125.41600201811417]
  We introduce SynthLLM, a scalable framework that transforms pre-training corpora into diverse, high-quality synthetic datasets.<n>Our approach achieves this by automatically extracting and recombining high-level concepts across multiple documents using a graph algorithm.
  arXiv  Detail & Related papers  (2025-03-25T11:07:12Z)
• LLM-TabLogic: Preserving Inter-Column Logical Relationships in Synthetic Tabular Data via Prompt-Guided Latent Diffusion [49.898152180805454]
  Synthetic datasets must maintain domain-specific logical consistency.<n>Existing generative models often overlook these inter-column relationships.<n>This study presents the first method to effectively preserve inter-column relationships without requiring domain knowledge.
  arXiv  Detail & Related papers  (2025-03-04T00:47:52Z)
• Zero-shot Meta-learning for Tabular Prediction Tasks with Adversarially Pre-trained Transformer [2.1677183904102257]
  We present an Adversarially Pre-trained Transformer (APT) that is able to perform zero-shot meta-learning on tabular prediction tasks without pre-training on any real-world dataset.<n>APT is pre-trained with adversarial synthetic data agents, who deliberately challenge the model with different synthetic datasets.<n>We show that our framework matches state-of-the-art performance on small classification tasks without filtering on dataset characteristics.
  arXiv  Detail & Related papers  (2025-02-06T23:58:11Z)
• TabDPT: Scaling Tabular Foundation Models on Real Data [20.00390825519329]
  We propose an approach to combine ICL-based retrieval with self supervised learning to train foundation models.<n>We show that incorporating real data during the pre-training phase can lead to significantly faster training and better generalization to unseen data.<n>Our resulting model, TabDPT, achieves top performance on both regression (CTR23) and classification (CC18) benchmarks.
  arXiv  Detail & Related papers  (2024-10-23T18:00:00Z)
• RelBench: A Benchmark for Deep Learning on Relational Databases [78.52438155603781]
  We present RelBench, a public benchmark for solving tasks over databases with graph neural networks. We use RelBench to conduct the first comprehensive study of Deep Learning infrastructure. RDL learns better whilst reducing human work needed by more than an order of magnitude.
  arXiv  Detail & Related papers  (2024-07-29T14:46:13Z)
• CTSyn: A Foundational Model for Cross Tabular Data Generation [9.568990880984813]
  Cross-Table Synthesizer (CTSyn) is a diffusion-based foundational model tailored for tabular data generation. CTSyn significantly outperforms existing table synthesizers in utility and diversity. It also uniquely enhances performances of downstream machine learning beyond what is achievable with real data.
  arXiv  Detail & Related papers  (2024-06-07T04:04:21Z)
• Federated Learning with Projected Trajectory Regularization [65.6266768678291]
  Federated learning enables joint training of machine learning models from distributed clients without sharing their local data. One key challenge in federated learning is to handle non-identically distributed data across the clients. We propose a novel federated learning framework with projected trajectory regularization (FedPTR) for tackling the data issue.
  arXiv  Detail & Related papers  (2023-12-22T02:12:08Z)
• Generating tabular datasets under differential privacy [0.0]
  We introduce Differential Privacy (DP) into the training process of deep neural networks. This creates a trade-off between the quality and privacy of the resulting data. We implement novel end-to-end models that leverage attention mechanisms.
  arXiv  Detail & Related papers  (2023-08-28T16:35:43Z)
• Generative Table Pre-training Empowers Models for Tabular Prediction [71.76829961276032]
  We propose TapTap, the first attempt that leverages table pre-training to empower models for tabular prediction. TapTap can generate high-quality synthetic tables to support various applications, including privacy protection, low resource regime, missing value imputation, and imbalanced classification. It can be easily combined with various backbone models, including LightGBM, Multilayer Perceptron (MLP) and Transformer.
  arXiv  Detail & Related papers  (2023-05-16T06:37:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.