# When Average Calibration Fails: Site-Conditional Federated Conformal Risk Control

arXiv: https://arxiv.org/abs/2606.20115 · HTML: https://mlsanity.com/paper/2606.20115/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 14 of its 18 references that are in the map.
- First posted on arXiv 18 Jun 2026.

## Builds on (most cited first, 14 of 14)

- [Communication-Efficient Learning of Deep Networks from Decentralized Data](https://mlsanity.com/paper/1602.05629/) (2016)
- [MONAI: An open-source framework for deep learning in healthcare](https://mlsanity.com/paper/2211.02701/) (2022)
- [Conformal Risk Control](https://mlsanity.com/paper/2208.02814/) (2022)
- [Federated Learning Enables Big Data for Rare Cancer Boundary Detection](https://mlsanity.com/paper/2204.10836/) (2022)
- [Federated Conformal Predictors for Distributed Uncertainty Quantification](https://mlsanity.com/paper/2305.17564/) (2023)
- [Conformal Prediction for Federated Uncertainty Quantification Under Label Shift](https://mlsanity.com/paper/2306.05131/) (2023)
- [Conformal Prediction for Image Segmentation Using Morphological Prediction Sets](https://mlsanity.com/paper/2503.05618/) (2025)
- [Certifiably Byzantine-Robust Federated Conformal Prediction](https://mlsanity.com/paper/2406.01960/) (2024)
- [Conformal Risk Control for Semantic Uncertainty Quantification in Computed Tomography](https://mlsanity.com/paper/2503.00136/) (2025)
- [Conditional Conformal Risk Adaptation](https://mlsanity.com/paper/2504.07611/) (2025)
- [FUNAvg: Federated Uncertainty Weighted Averaging for Datasets with Diverse Labels](https://mlsanity.com/paper/2407.07488/) (2024)
- [Efficient Federated Conformal Prediction with Group-Conditional Guarantee](https://mlsanity.com/paper/2603.14198/) (2026)
- [Conformalized Neural Networks for Federated Uncertainty Quantification under Dual Heterogeneity](https://mlsanity.com/paper/2602.23296/) (2026)
- [FedStein: Enhancing Multi-Domain Federated Learning Through James-Stein Estimator](https://mlsanity.com/paper/2410.03499/) (2024)

## Abstract (arXiv)

Conformal risk control (CRC) provides distribution-free segmentation guarantees by calibrating a prediction-set threshold on held-out data. In federated deployments, the standard approach pools calibration scores into a single threshold. We quantify, on real multi-institutional brain tumor data (FeTS-2022, 1,251 subjects, 20 institutions), a critical failure: naive pooled CRC protects the average hospital but violates coverage at 40% of individual institutions, with the worst site exceeding the target false-negative rate by 7.8 percentage points. We trace this failure to a hidden design choice: the aggregation weights implicitly determine whose coverage is protected. Sample-size weighting optimizes patient-level validity but can sacrifice institution-level reliability; equal-site weighting improves institution-level reliability on this benchmark at comparable efficiency, using only a single scalar per site. We propose risk-curve shrinkage as a principled mechanism: each site transmits its empirical risk curve (G scalars) and a single hyperparameter n0 smoothly interpolates between site-specific local calibration and sample-size-weighted pooled calibration. Leave-one-site-out sensitivity analysis identifies n0=19, achieving 2.7/20 violations at 2.0x stretch. Direct Lagrangian budget optimization fails by concentrating risk on vulnerable hospitals; the finite-sample correction term is essential: removing it triples violations. No patient-level images, masks, or per-volume scores leave any site.