الملخص
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 $\mathrm{ECE}_{15}$ 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.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Liang, W., Zheng, L. N., Yue, L., Zhang, W. E., Guo, M., Maennel, O., & Chen, W. (2026). Signal-Routed Temperature Scaling: Low-Capacity Risk-Conditioned Calibration for Small Validation Budgets. https://omanscience.com/ar/articles/signal-routed-temperature-scaling-low-capacity-risk-conditioned-calibration-for-small-validation-budgets
MLA 9
Liang, Wenhao, et al. "Signal-Routed Temperature Scaling: Low-Capacity Risk-Conditioned Calibration for Small Validation Budgets." https://omanscience.com/ar/articles/signal-routed-temperature-scaling-low-capacity-risk-conditioned-calibration-for-small-validation-budgets.
شيكاغو (المؤلف–التاريخ)
Liang, Wenhao, Liangwei Nathan Zheng, Lin Yue, Wei Emma Zhang, Mingyu Guo, Olaf Maennel, and Weitong Chen. 2026. "Signal-Routed Temperature Scaling: Low-Capacity Risk-Conditioned Calibration for Small Validation Budgets." https://omanscience.com/ar/articles/signal-routed-temperature-scaling-low-capacity-risk-conditioned-calibration-for-small-validation-budgets.
هارفارد
Liang, W., Zheng, L. N., Yue, L., Zhang, W. E., Guo, M., Maennel, O. and Chen, W. (2026) 'Signal-Routed Temperature Scaling: Low-Capacity Risk-Conditioned Calibration for Small Validation Budgets', Available at: https://omanscience.com/ar/articles/signal-routed-temperature-scaling-low-capacity-risk-conditioned-calibration-for-small-validation-budgets.
فانكوفر
Liang W, Zheng LN, Yue L, Zhang WE, Guo M, Maennel O, et al. Signal-Routed Temperature Scaling: Low-Capacity Risk-Conditioned Calibration for Small Validation Budgets. https://omanscience.com/ar/articles/signal-routed-temperature-scaling-low-capacity-risk-conditioned-calibration-for-small-validation-budgets
IEEE
W. Liang, L. N. Zheng, L. Yue, W. E. Zhang, M. Guo, O. Maennel, and W. Chen, "Signal-Routed Temperature Scaling: Low-Capacity Risk-Conditioned Calibration for Small Validation Budgets," https://omanscience.com/ar/articles/signal-routed-temperature-scaling-low-capacity-risk-conditioned-calibration-for-small-validation-budgets.