Research map / Architectures and efficiency / RNNs and normalization
Revisiting Deep Learning Models for Tabular Data
Yury Gorishniy, Ivan Rubachev, Valentin Khrulkov, Artem Babenko
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add feature biases to Feature Tokenizer (ablation) | FT-Transformer · Covertype | accuracy: 0.964 -> 0.970 | Improved |
| add feature biases to Feature Tokenizer (ablation) | FT-Transformer · Year | RMSE: 8.843 -> 8.855 | Hurt |
| vanilla Transformer backbone with [CLS] and feature biases instead of AutoInt design (ablation) | attention-based tabular model · Covertype | accuracy: 0.934 -> 0.970 | Improved |
| FT-Transformer instead of ResNet | tabular DL · 11 tabular datasets | average rank: 3.3 (1.8) -> 1.8 (1.2) | Improved |
| tuned GBDT ensembles instead of FT-Transformer ensembles | tabular models · California Housing | RMSE: 0.448 -> 0.423 | Improved |
| FT-Transformer instead of GBDT | tabular models · Helena | accuracy: 0.388 -> 0.398 | Improved |
| tune hyperparameters of simple baselines (MLP, ResNet) | MLP/ResNet · 11 tabular datasets | rank | Improved |
| ensembling | FT-Transformer · California Housing | RMSE: 0.459 -> 0.448 | Improved |
| attention-map feature importance instead of Integrated Gradients | FT-Transformer · Year | rank correlation with permutation test: 0.50 (0.03) -> 0.92 (0.01) | Improved |
Closest papers by meaning
- TabR: Tabular Deep Learning Meets Nearest Neighbors (ICLR 2024 · similarity 0.92)
- TabM: Advancing tabular deep learning with parameter-efficient ensembling (ICLR 2025 · similarity 0.92)
- TabReD: Analyzing Pitfalls and Filling the Gaps in Tabular Deep Learning Benchmarks (ICLR 2025 · similarity 0.90)
- TabSTAR: A Tabular Foundation Model for Tabular Data with Text Fields (NeurIPS 2025 · similarity 0.90)
- Retrieval & Fine-Tuning for In-Context Tabular Models (NeurIPS 2024 · similarity 0.89)
- TabArena: A Living Benchmark for Machine Learning on Tabular Data (NeurIPS 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
