---
title: "Tabular foundation models are not a shortcut to calibrated clinical risk prediction: a pre-registered simulation against the true risk"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2610.39065v1"
permalink: "https://feed.astralor.com/items/itm_55cb505cee8a7ffb97a7a4ee"
published: "2026-10-28T23:51:50.000Z"
score: 53
channel: "ai"
tags: ["RAG","Tabular","Whether","TabPFN-3","TabICLv2","Mitra-v2"]
---

# Tabular foundation models are not a shortcut to calibrated clinical risk prediction: a pre-registered simulation against the true risk

## At a glance

Tabular foundation models are pretrained networks that predict risk from a table without fitting a model. Whether their risks are correct (calibrated) has been judged on real data, where true risk is unknown. In a pre-r…

## Key points

- Tabular foundation models are pretrained networks that predict risk from a table without fitting a model.
- Whether their risks are correct (calibrated) has been judged on real data, where true risk is unknown.
- In a pre-registered simulation with known true risks, we compared TabPFN-3.5, TabICLv2 and Mitra-v2 with Firth logistic regression with intercept correction (FLIC) in 58 scenarios and 1,000 random designs, then on four …

## Full text

Tabular foundation models are pretrained networks that predict risk from a table without fitting a model. Whether their risks are correct (calibrated) has been judged on real data, where true risk is unknown. In a pre-registered simulation with known true risks, we compared TabPFN-3.5, TabICLv2 and Mitra-v2 with Firth logistic regression with intercept correction (FLIC) in 58 scenarios and 1,000 random designs, then on four clinical data sets. Foundation models were not a shortcut to calibration. Their calibration slope matched FLIC's within the pre-registered margin in 36 or 37 of 48 core scenarios and was better only at n = 200 with a 3% event rate. Even 3,000 patients were often not enough: with FLIC, TabPFN-3.5 or Mitra-v2, 63%-83% of samples gave risks close to the truth in average, intercept and slope at event rates of 10% and 30%, and 35%-50% at 3%. TabICLv2 did worse (8%-39% and 1%-3%). Across the random designs, foundation models had a lower true calibration error than FLIC beyond the margin in 25%-38% and a higher one in none, though small losses were predicted at n = 3000. On real data, no model resisted a change in event rate between populations; recalibration on 500 local patients largely repaired it.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_55cb505cee8a7ffb97a7a4ee