Sharp Class-Count Bounds for Calibration Reuse in Temperature-Scaled Conformal Prediction

How much coverage can a conformal classifier lose when it uses the same observations to select a temperature and calibrate its prediction sets? For nonrandomized least-ambiguous classification with a fixed K-class softmax model, the temperature-and-threshold family has growth function at most m + 1 + (K − 2)m(m − 1) and VC dimension asymptotic to log₂ K. One fixed classifier can realize K − 1 independent score coordinates at prescribed temperatures. This gives an exact beta order-statistic law for the coverage of an adversarial calibration-only selector. At interior finite-rank coverage levels, the worst-selector marginal deficit has order min{1, √(log K / n)}, where n is the calibration sample size. The lower construction allows bounded logits and temperatures in any nonempty open interval of positive values. A three-class example compares adversarial selection with likelihood tuning. The results characterize worst-case calibration reuse for unmodified LAC sets.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-19
DOI
https://doi.org/10.5281/zenodo.24374512
Primary Topic
Machine Learning and Algorithms
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Sharp Class-Count Bounds for Calibration Reuse in Temperature-Scaled Conformal Prediction

S. M. Abdulla Al Mamun
Zenodo (CERN European Organization for Nuclear Research)
Machine Learning and Algorithms
preprint

Sharp Class-Count Bounds for Calibration Reuse in Temperature-Scaled Conformal Prediction

S. M. Abdulla Al Mamun
preprint en

Abstract

How much coverage can a conformal classifier lose when it uses the same observations to select a temperature and calibrate its prediction sets? For nonrandomized least-ambiguous classification with a fixed K-class softmax model, the temperature-and-threshold family has growth function at most m + 1 + (K − 2)m(m − 1) and VC dimension asymptotic to log₂ K. One fixed classifier can realize K − 1 independent score coordinates at prescribed temperatures. This gives an exact beta order-statistic law for the coverage of an adversarial calibration-only selector. At interior finite-rank coverage levels, the worst-selector marginal deficit has order min{1, √(log K / n)}, where n is the calibration sample size. The lower construction allows bounded logits and temperatures in any nonempty open interval of positive values. A three-class example compares adversarial selection with likelihood tuning. The results characterize worst-case calibration reuse for unmodified LAC sets.

Zenodo (CERN European Organization for Nuclear Research)
University of Manchester (GB)
Machine Learning and Algorithms
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon