Prior-data fitted networks for variable selection
stat.ME, stat.ML
Submitted: 2026-11-04
Updated: 2026-11-04
Code: https://github.com/omelikechi/varselpfn
Terminology
Sources
- DCD-PFN: A Decoupling-Aware Foundation Model for Causal Discovery
- TabFM: A Zero-Shot Foundation Model for Tabular Data
- SelectPFN: Amortized Bayesian Variable Selection for Sparse Nonlinear Regression with Prior-Data Fitted Networks
- Engineering Regression Without Real-Data Training: Domain Adaptation for Tabular Foundation Models Using Multi-Dataset Embeddings
- LinearPFN: Amortized Variable Selection for Linear Models with Interactions
- Layer Normalization
- Just add noise: Debiasing tree-based variable importance in mixed data
- nanoTabPFN: A Lightweight and Educational Reimplementation of TabPFN
Related papers
- Learning Robust Treatment Rules for Censored Data
- Neural Estimation of Summary Statistics for Simulation-Based Inference with Non-Asymptotic Guarantees
- Spectral divide-and-conquer MCMC for long stationary time series
- Improving the Efficiency of Subgroup Analysis in Randomized Controlled Trials with TMLE
- Personalized Federated Vector Autoregression with Personalization Diversity
- Doubly Robust and Efficient Calibration of Prediction Sets for Right-Censored Time-to-Event Outcomes