Valid Feature-Level Inference for Tabular Foundation Models via the Conditional Randomization Test
Mohamed Salem · 2026
Read Valid Feature-Level Inference for Tabular Foundation Models via the Conditional Randomization Test by Mohamed Salem (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-06609.txt.
Description
Salem presents a practical feature-level hypothesis testing approach joining the Conditional Shuffle Test with TabPFN, a tabular-data chance foundation model. Modern models are highly rich but difficult to analyze by stats. Black-box guessers rarely give valid tests or p-values for whether single features carry information about a target variable. The final procedure gives finite-sample valid p-values for conditional relevance of inputs, even under nonlinear and linked settings, without new training or set assumptions.
Catalog record
| CREATOR | Mohamed Salem |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25243 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.06609