# Signal-Routed Temperature Scaling: Low-Capacity Risk-Conditioned Calibration for Small Validation Budgets

arXiv: https://arxiv.org/abs/2609.38936 · HTML: https://mlsanity.com/paper/2609.38936/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 26 of its 29 references that are in the map.
- First posted on arXiv 30 Sep 2026.

## Builds on (most cited first, 26 of 26)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Rethinking the Inception Architecture for Computer Vision](https://mlsanity.com/paper/1512.00567/) (2015)
- [mixup: Beyond Empirical Risk Minimization](https://mlsanity.com/paper/1710.09412/) (2017)
- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [Training data-efficient image transformers & distillation through attention](https://mlsanity.com/paper/2012.12877/) (2020)
- [CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features](https://mlsanity.com/paper/1905.04899/) (2019)
- [Benchmarking Neural Network Robustness to Common Corruptions and Perturbations](https://mlsanity.com/paper/1903.12261/) (2019)
- [RandAugment: Practical automated data augmentation with a reduced search space](https://mlsanity.com/paper/1909.13719/) (2019)
- [Designing Network Design Spaces](https://mlsanity.com/paper/2003.13678/) (2020)
- [Selective Classification for Deep Neural Networks](https://mlsanity.com/paper/1705.08500/) (2017)
- [Measuring Calibration in Deep Learning](https://mlsanity.com/paper/1904.01685/) (2019)
- [Beyond temperature scaling: Obtaining well-calibrated multiclass probabilities with Dirichlet calibration](https://mlsanity.com/paper/1910.12656/) (2019)
- [Mix-n-Match: Ensemble and Compositional Methods for Uncertainty Calibration in Deep Learning](https://mlsanity.com/paper/2003.07329/) (2020)
- [Bias-Reduced Uncertainty Estimation for Deep Neural Classifiers](https://mlsanity.com/paper/1805.08206/) (2018)
- [Local Temperature Scaling for Probability Calibration](https://mlsanity.com/paper/2008.05105/) (2020)
- [Smooth ECE: Principled Reliability Diagrams via Kernel Smoothing](https://mlsanity.com/paper/2309.12236/) (2023)
- [Sample-dependent Adaptive Temperature Scaling for Improved Calibration](https://mlsanity.com/paper/2207.06211/) (2022)
- [Meta-Cal: Well-controlled Post-hoc Calibration by Ranking](https://mlsanity.com/paper/2105.04290/) (2021)
- [Confidence Calibration of Classifiers with Many Classes](https://mlsanity.com/paper/2411.02988/) (2024)
- [Beyond Probability Partitions: Calibrating Neural Networks with Semantic Aware Grouping](https://mlsanity.com/paper/2306.04985/) (2023)
- [Sample Margin-Aware Recalibration of Temperature Scaling](https://mlsanity.com/paper/2506.23492/) (2025)
- [CalArena: A Large-Scale Post-Hoc Calibration Benchmark](https://mlsanity.com/paper/2605.30188/) (2026)
- [Quantile Adaptive Temperature Scaling for Confidence Calibration](https://mlsanity.com/paper/2606.21749/) (2026)
- [Towards Precise Prediction Uncertainty in GNNs: Refining GNNs with Topology-grouping Strategy](https://mlsanity.com/paper/2412.14223/) (2024)

## Abstract (arXiv)

When a classifier is recalibrated from only a few thousand held-out examples, the capacity of the calibration map becomes a statistical design choice rather than a purely architectural one: a scalar map can underfit structured residual miscalibration, while a highly adaptive map can be hard to estimate reliably from so small a split. We disentangle the calibration objective from adaptive capacity and propose signal-routed temperature scaling (SRTS-BCE), a 10-parameter, argmax-preserving calibrator that cross-fits a correctness-risk score over six logit statistics and fits one top-label-BCE temperature per K=3 risk groups, recovering TvA-TS as its K=1 limit. On fine-tuned CIFAR-100 / ViT-B/16, SRTS-BCE reduces ECE15 from 1.65 (scalar TvA-TS) to 0.96, matching the higher-capacity SMART+BCE head (0.95) at the full calibration budget. The two regimes separate as the budget shrinks: at n=250 SRTS-BCE beats SMART+BCE on all three CIFAR-100 backbones (the seed-to-draw hierarchical interval excludes zero), whereas the flagship comparison against the scalar remains directional. A protocol-frozen Tiny-ImageNet follow-up reproduces the small-budget separation and exhibits a budget-dependent ranking reversal on Swin-T; matched routing and map controls show that the effect is tied neither to the learned router nor to discrete grouping. Together the results identify post-hoc calibrator capacity as a finite-sample design choice whose preferred level shifts with the amount of available calibration data.