TabICL: A Tabular Foundation Model for In-Context Learning on Large Data

• URL: http://arxiv.org/abs/2502.05564v1
• Date: Sat, 08 Feb 2025 13:25:04 GMT
• Title: TabICL: A Tabular Foundation Model for In-Context Learning on Large Data
• Authors: Jingang Qu, David Holzmüller, Gaël Varoquaux, Marine Le Morvan,
• Abstract summary: We introduce TabICL, a foundation model for classification pretrained on synthetic datasets with up to 60K samples. It is on par with TabPFNv2 while being systematically faster (up to 10 times) and significantly outperforms all other approaches. On 56 datasets with over 10K samples, TabICL surpasses both TabPFNv2 and CatBoost, demonstrating the potential of ICL for large data.
• Score: 15.08819125687632
• License:
• Abstract: The long-standing dominance of gradient-boosted decision trees on tabular data is currently challenged by tabular foundation models using In-Context Learning (ICL): setting the training data as context for the test data and predicting in a single forward pass without parameter updates. While the very recent TabPFNv2 foundation model (2025) excels on tables with up to 10K samples, its alternating column- and row-wise attentions make handling large training sets computationally prohibitive. So, can ICL be effectively scaled and deliver a benefit for larger tables? We introduce TabICL, a tabular foundation model for classification, pretrained on synthetic datasets with up to 60K samples and capable of handling 500K samples on affordable resources. This is enabled by a novel two-stage architecture: a column-then-row attention mechanism to build fixed-dimensional embeddings of rows, followed by a transformer for efficient ICL. Across 200 classification datasets from the TALENT benchmark, TabICL is on par with TabPFNv2 while being systematically faster (up to 10 times), and significantly outperforms all other approaches. On 56 datasets with over 10K samples, TabICL surpasses both TabPFNv2 and CatBoost, demonstrating the potential of ICL for large data.

Related papers

• Transformers Boost the Performance of Decision Trees on Tabular Data across Sample Sizes [135.68092471784516]
  We propose a simple and lightweight approach for fusing large language models and gradient-boosted decision trees. We name our fusion methods LLM-Boost and PFN-Boost, respectively. We demonstrate state-of-the-art performance against numerous baselines and ensembling algorithms.
  arXiv  Detail & Related papers  (2025-02-04T19:30:41Z)
• TabDPT: Scaling Tabular Foundation Models [20.00390825519329]
  We show how to harness the power of real data to improve performance and generalization. Our model achieves state-of-the-art performance on the CC18 (classification) and CTR23 (regression) benchmarks. TabDPT also demonstrates strong scaling as both model size and amount of available data increase.
  arXiv  Detail & Related papers  (2024-10-23T18:00:00Z)
• LaTable: Towards Large Tabular Models [63.995130144110156]
  Tabular generative foundation models are hard to build due to the heterogeneous feature spaces of different datasets. LaTable is a novel diffusion model that addresses these challenges and can be trained across different datasets. We find that LaTable outperforms baselines on in-distribution generation, and that finetuning LaTable can generate out-of-distribution datasets better with fewer samples.
  arXiv  Detail & Related papers  (2024-06-25T16:03:50Z)
• Mixture of In-Context Prompters for Tabular PFNs [33.76194735049027]
  MIXTUREPFN is the Condorcet winner across 36 diverse datasets against 19 strong deep learning and tree-based baselines. It achieves the highest mean rank among Top-10 aforementioned algorithms with statistical significance.
  arXiv  Detail & Related papers  (2024-05-25T09:47:59Z)
• TuneTables: Context Optimization for Scalable Prior-Data Fitted Networks [90.00817095558094]
  Prior-data fitted networks (PFNs) make use of pretraining and in-context learning to achieve strong performance on new tasks in a single forward pass. We introduce TuneTables, a parameter-efficient fine-tuning strategy for PFNs that compresses large datasets into a smaller learned context. We show that TuneTables can be used as an interpretability tool and can even be used to mitigate biases by optimizing a fairness objective.
  arXiv  Detail & Related papers  (2024-02-17T00:02:23Z)
• In-Context Data Distillation with TabPFN [11.553950697974825]
  In-context data distillation (ICD) is a novel methodology that effectively eliminates these constraints by optimizing TabPFN's context. ICD efficiently enables TabPFN to handle significantly larger datasets with a fixed memory budget, improving TabPFN's quadratic memory complexity but at the cost of a linear number of tuning steps.
  arXiv  Detail & Related papers  (2024-02-10T15:23:45Z)
• Rethinking Pre-Training in Tabular Data: A Neighborhood Embedding Perspective [71.45945607871715]
  We propose Tabular data Pre-Training via Meta-representation (TabPTM) The core idea is to embed data instances into a shared feature space, where each instance is represented by its distance to a fixed number of nearest neighbors and their labels. Extensive experiments on 101 datasets confirm TabPTM's effectiveness in both classification and regression tasks, with and without fine-tuning.
  arXiv  Detail & Related papers  (2023-10-31T18:03:54Z)
• 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)
• TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second [48.87527918630822]
  We present TabPFN, a trained Transformer that can do supervised classification for small datasets in less than a second. TabPFN performs in-context learning (ICL), it learns to make predictions using sequences of labeled examples. We show that our method clearly outperforms boosted trees and performs on par with complex state-of-the-art AutoML systems with up to 230$times$ speedup.
  arXiv  Detail & Related papers  (2022-07-05T07:17:43Z)
• Scientific evidence extraction [0.0]
  We propose a new dataset, Tables One Million (PubTables-1M), and a new class of metric, PubMed grid table similarity (GriTS) PubTables-1M is nearly twice as large as the previous largest comparable dataset. We show that object detection models trained on PubTables-1M produce excellent results out-of-the-box for all three tasks of detection, structure recognition, and functional analysis.
  arXiv  Detail & Related papers  (2021-09-30T19:42:07Z)

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.