# Valid Feature-Level Inference for Tabular Foundation Models via the Conditional Randomization Test

arXiv: https://arxiv.org/abs/2603.06609 · HTML: https://mlsanity.com/paper/2603.06609/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 8 of its 22 references that are in the map.
- First posted on arXiv 19 Feb 2026.

## Builds on (most cited first, 8 of 8)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- ["Why Should I Trust You?": Explaining the Predictions of Any Classifier](https://mlsanity.com/paper/1602.04938/) (2016)
- [TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Second](https://mlsanity.com/paper/2207.01848/) (2022)
- [Panning for Gold: Model-X Knockoffs for High-dimensional Controlled Variable Selection](https://mlsanity.com/paper/1610.02351/) (2016)
- [Controlling the false discovery rate via knockoffs](https://mlsanity.com/paper/1404.5609/) (2014)
- [On the power of conditional independence testing under model-X](https://mlsanity.com/paper/2005.05506/) (2020)

## Built on by (most cited first, 1 of 1)

- [TabPFN-3: Technical Report](https://mlsanity.com/paper/2605.13986/) (2026)

## Abstract (arXiv)

Modern machine learning models are highly expressive but notoriously difficult to analyze statistically. In particular, while black-box predictors can achieve strong empirical performance, they rarely provide valid hypothesis tests or p-values for assessing whether individual features contain information about a target variable. This article presents a practical approach to feature-level hypothesis testing that combines the Conditional Randomization Test (CRT) with TabPFN, a probabilistic foundation model for tabular data. The resulting procedure yields finite-sample valid p-values for conditional feature relevance, even in nonlinear and correlated settings, without requiring model retraining or parametric assumptions.