Aggregate Tabular Benchmarking on Aggregate
1Avg RankTabII
Evaluation results
| Method | Links | ||
|---|---|---|---|
| TabII2026.01 | 1 | 0 | |
| MLP-PLRBackbone=MLP, Numerical Feature Embedding=Piecewise Linear2022.03 | 3 | — | |
| ResNet-PLRBackbone=ResNet, Numerical Feature Embedding=Piecewise Linear2022.03 | 3.2 | — | |
| CatBoostBackbone=GBDT2022.03 | 3.6 | — | |
| Transformer-T-LRBackbone=Transformer, Numerical Feature Embedding=Periodic2022.03 | 3.7 | — | |
| Transformer-PLRBackbone=Transformer, Numerical Feature Embedding=Piecewise Linear2022.03 | 3.9 | — | |
| ResNet-T-LRBackbone=ResNet, Numerical Feature Embedding=Periodic2022.03 | 4.1 | — | |
| Transformer-Q-LRBackbone=Transformer, Numerical Feature Embedding=Quantile2022.03 | 4.4 | — | |
| XGBoostBackbone=GBDT2022.03 | 4.6 | — | |
| ResNet-Q-LRBackbone=ResNet, Numerical Feature Embedding=Quantile2022.03 | 4.7 | — | |
| MLP-Q-LRBackbone=MLP, Numerical Feature Embedding=Quantile2022.03 | 5.1 | — | |
| MLP-T-LRBackbone=MLP, Numerical Feature Embedding=Periodic2022.03 | 5.1 | — | |
| TabPFN-v2ICL Setup=add test with pseudo label2026.01 | 5.19 | 4.54 | |
| Transformer-LRBackbone=Transformer, Numerical Feature Embedding=Linear2022.03 | 5.2 | — | |
| MLP-LRBackbone=MLP, Numerical Feature Embedding=Linear2022.03 | 5.5 | — | |
| ResNet-LRBackbone=ResNet, Numerical Feature Embedding=Linear2022.03 | 5.6 | — | |
| FT-TransInference Columns=Train set only2026.01 | 5.69 | 2.25 | |
| TabPFN-v2ICL Setup=add test without label2026.01 | 5.69 | 3.07 | |
| Transformer-LBackbone=Transformer2022.03 | 5.9 | — | |
| ResNetBackbone=ResNet2022.03 | 6.7 | — | |
| ResNetInference Columns=Train set only2026.01 | 6.75 | 3.28 | |
| CatBoostInference Columns=Train set only2026.01 | 7.44 | 3.94 | |
| MLPInference Columns=Train set only2026.01 | 8.06 | 2.69 | |
| TransTabInference Columns=All columns2026.01 | 8.25 | 5.44 | |
| MLPBackbone=MLP2022.03 | 8.5 | — | |
| SCARFInference Columns=Train set only2026.01 | 8.56 | 2.76 | |
| TabPFN-v2ICL Setup=with train set only2026.01 | 9.38 | 4.37 | |
| XGBoostInference Columns=Train set only2026.01 | 9.75 | 5.04 | |
| LRInference Columns=Train set only2026.01 | 9.88 | 3.02 | |
| MediTabModel Type=LLM-assisted2026.01 | 10.75 | 4.4 | |
| TabNetInference Columns=Train set only2026.01 | 11.75 | 1.67 | |
| TabLLMModel Type=LLM-assisted2026.01 | 13.31 | 2.34 | |
| FT-Trans*Inference Columns=All columns, Embedding=Linear2026.01 | 14.56 | 2.8 |