Orthogonal Gradient Descent Improves Neural Calibration

• URL: http://arxiv.org/abs/2506.04487v2
• Date: Sat, 28 Jun 2025 16:19:43 GMT
• Title: Orthogonal Gradient Descent Improves Neural Calibration
• Authors: C. Evans Hedges,
• Abstract summary: OnAR-10 with 10% labeled data, $perp$Grad matches SGD in accuracy but yields consistently improved calibration metrics.<n>These benefits persist under input corruption (CIFAR-10C) and extended training, where $perp$Grad models degrade more gracefully than SGD-trained counterparts.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We provide evidence that orthogonalizing gradients during training improves model calibration without sacrificing accuracy. On CIFAR-10 with 10\% labeled data, $\perp$Grad matches SGD in accuracy but yields consistently improved calibration metrics such as lower test loss, reduced softmax overconfidence, and higher predictive entropy. These benefits persist under input corruption (CIFAR-10C) and extended training, where $\perp$Grad models degrade more gracefully than SGD-trained counterparts. $\perp$Grad is optimizer-agnostic, incurs minimal overhead, and works well with post-hoc calibration techniques like temperature scaling. Theoretically, we prove convergence of a simplified version of $\perp$Grad under mild assumptions and characterize its stationary points in positive homogeneous networks: $\perp$Grad converges to solutions where further loss reduction requires confidence scaling rather than decision boundary improvement. Code for this paper can be found at: https://github.com/evanshedges2/orthograd\_improves\_calibration.

Related papers

• Theoretical Framework for Tempered Fractional Gradient Descent: Application to Breast Cancer Classification [0.0]
  This paper introduces Fractional Gradient Descent (TFGD), a novel optimization framework that synergizes fractional calculus with exponential tempering to enhance gradient-based learning.<n>TFGD addresses limitations by incorporating a tempered memory mechanism, where historical gradients are weighted by fractional coefficients $|w_j| = binomalphaj$ and exponentially decayed via a tempering parameter $lambda$.<n> Empirical validation on the Breast Cancer Wisconsin dataset demonstrates TFGD's superiority, achieving 98.25% test accuracy (vs. 92.11% for SGD) and 2$times$ faster convergence.
  arXiv  Detail & Related papers  (2025-04-26T08:26:34Z)
• Gradient Normalization Provably Benefits Nonconvex SGD under Heavy-Tailed Noise [60.92029979853314]
  We investigate the roles of gradient normalization and clipping in ensuring the convergence of Gradient Descent (SGD) under heavy-tailed noise. Our work provides the first theoretical evidence demonstrating the benefits of gradient normalization in SGD under heavy-tailed noise. We introduce an accelerated SGD variant incorporating gradient normalization and clipping, further enhancing convergence rates under heavy-tailed noise.
  arXiv  Detail & Related papers  (2024-10-21T22:40:42Z)
• Feature Clipping for Uncertainty Calibration [24.465567005078135]
  Modern deep neural networks (DNNs) often suffer from overconfidence, leading to miscalibration. We propose a novel post-hoc calibration method called feature clipping (FC) to address this issue. FC involves clipping feature values to a specified threshold, effectively increasing entropy in high calibration error samples.
  arXiv  Detail & Related papers  (2024-10-16T06:44:35Z)
• Novel Gradient Sparsification Algorithm via Bayesian Inference [27.246907664193156]
  This paper proposes a novel sparsification algorithm called regularized Top-$k$ (RegTop-$k$) that controls the learning rate scaling of error accumulation. Numerical experiments with ResNet-18 on CIFAR-10 show that RegTop-$k$ achieves about $8%$ higher accuracy than standard Top-$k$.
  arXiv  Detail & Related papers  (2024-09-23T10:42:34Z)
• Orthogonal Causal Calibration [55.28164682911196]
  We develop general algorithms for reducing the task of causal calibration to that of calibrating a standard (non-causal) predictive model.<n>Our results are exceedingly general, showing that essentially any existing calibration algorithm can be used in causal settings.
  arXiv  Detail & Related papers  (2024-06-04T03:35:25Z)
• On Convergence of Incremental Gradient for Non-Convex Smooth Functions [63.51187646914962]
  In machine learning and network optimization, algorithms like shuffle SGD are popular due to minimizing the number of misses and good cache. This paper delves into the convergence properties SGD algorithms with arbitrary data ordering.
  arXiv  Detail & Related papers  (2023-05-30T17:47:27Z)
• Bridging Precision and Confidence: A Train-Time Loss for Calibrating Object Detection [58.789823426981044]
  We propose a novel auxiliary loss formulation that aims to align the class confidence of bounding boxes with the accurateness of predictions. Our results reveal that our train-time loss surpasses strong calibration baselines in reducing calibration error for both in and out-domain scenarios.
  arXiv  Detail & Related papers  (2023-03-25T08:56:21Z)
• Variance-reduced Clipping for Non-convex Optimization [24.765794811146144]
  Gradient clipping is a technique used in deep learning applications such as large-scale language modeling. Recent experimental training have a fairly special behavior in that it mitigates order complexity.
  arXiv  Detail & Related papers  (2023-03-02T00:57:38Z)
• AdaFocal: Calibration-aware Adaptive Focal Loss [8.998525155518836]
  Training with focal loss leads to better calibration than cross-entropy. We propose a calibration-aware adaptive focal loss called AdaFocal.
  arXiv  Detail & Related papers  (2022-11-21T20:19:24Z)
• Improved techniques for deterministic l2 robustness [63.34032156196848]
  Training convolutional neural networks (CNNs) with a strict 1-Lipschitz constraint under the $l_2$ norm is useful for adversarial robustness, interpretable gradients and stable training. We introduce a procedure to certify robustness of 1-Lipschitz CNNs by replacing the last linear layer with a 1-hidden layer. We significantly advance the state-of-the-art for standard and provable robust accuracies on CIFAR-10 and CIFAR-100.
  arXiv  Detail & Related papers  (2022-11-15T19:10:12Z)
• A Closer Look at the Calibration of Differentially Private Learners [33.715727551832785]
  We study the calibration of classifiers trained with differentially private descent gradient (DP-SGD) Our analysis identifies per-example gradient clipping in DP-SGD as a major cause of miscalibration. We show that differentially private variants of post-processing calibration methods such as temperature scaling and Platt scaling are surprisingly effective.
  arXiv  Detail & Related papers  (2022-10-15T10:16:18Z)
• Sample-dependent Adaptive Temperature Scaling for Improved Calibration [95.7477042886242]
  Post-hoc approach to compensate for neural networks being wrong is to perform temperature scaling. We propose to predict a different temperature value for each input, allowing us to adjust the mismatch between confidence and accuracy. We test our method on the ResNet50 and WideResNet28-10 architectures using the CIFAR10/100 and Tiny-ImageNet datasets.
  arXiv  Detail & Related papers  (2022-07-13T14:13:49Z)
• When are Iterative Gaussian Processes Reliably Accurate? [38.523693700243975]
  Lanczos decompositions have achieved scalable Gaussian process inference with highly accurate point predictions. We investigate CG tolerance, preconditioner rank, and Lanczos decomposition rank. We show that LGS-BFB is a compelling for Iterative GPs, achieving convergence with fewer updates.
  arXiv  Detail & Related papers  (2021-12-31T00:02:18Z)
• High-probability Bounds for Non-Convex Stochastic Optimization with Heavy Tails [55.561406656549686]
  We consider non- Hilbert optimization using first-order algorithms for which the gradient estimates may have tails. We show that a combination of gradient, momentum, and normalized gradient descent convergence to critical points in high-probability with best-known iteration for smooth losses.
  arXiv  Detail & Related papers  (2021-06-28T00:17:01Z)
• Parameterized Temperature Scaling for Boosting the Expressive Power in Post-Hoc Uncertainty Calibration [57.568461777747515]
  We introduce a novel calibration method, Parametrized Temperature Scaling (PTS) We demonstrate that the performance of accuracy-preserving state-of-the-art post-hoc calibrators is limited by their intrinsic expressive power. We show with extensive experiments that our novel accuracy-preserving approach consistently outperforms existing algorithms across a large number of model architectures, datasets and metrics.
  arXiv  Detail & Related papers  (2021-02-24T10:18:30Z)
• Extrapolation for Large-batch Training in Deep Learning [72.61259487233214]
  We show that a host of variations can be covered in a unified framework that we propose. We prove the convergence of this novel scheme and rigorously evaluate its empirical performance on ResNet, LSTM, and Transformer.
  arXiv  Detail & Related papers  (2020-06-10T08:22:41Z)
• Towards Better Understanding of Adaptive Gradient Algorithms in Generative Adversarial Nets [71.05306664267832]
  Adaptive algorithms perform gradient updates using the history of gradients and are ubiquitous in training deep neural networks. In this paper we analyze a variant of OptimisticOA algorithm for nonconcave minmax problems. Our experiments show that adaptive GAN non-adaptive gradient algorithms can be observed empirically.
  arXiv  Detail & Related papers  (2019-12-26T22:10:10Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.