Gradient-Descent Steps to Success over Mean Accuracy: A Paradigm Shift for ML

• URL: http://arxiv.org/abs/2606.22053v1
• Date: Sat, 20 Jun 2026 14:10:49 GMT
• Title: Gradient-Descent Steps to Success over Mean Accuracy: A Paradigm Shift for ML
• Abstract summary: We propose a paradigm shift towards evaluating performance based on computational effort-explicitly defined here as the total number of gradient descent steps required to reach an acceptable level of accuracy with high probability.<n>We demonstrate that minimising this effort acts as a novel form of Automatic Machine Learning (AutoML)
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Traditional evaluation of machine learning (ML) models typically focuses on achieving the maximum possible accuracy irrespective of the computational cost. In this article, we propose a paradigm shift towards evaluating performance based on computational effort-explicitly defined here as the total number of gradient descent steps required to reach an acceptable level of accuracy with high probability. Building upon the concept of computational effort originally introduced by Koza for Genetic Programming, we extend this metric to any ML model trained via gradient descent. Furthermore, we demonstrate that minimising this effort acts as a novel form of Automatic Machine Learning (AutoML). By evaluating it across 11 diverse ML models and five standard classification datasets, we uncover significant insights into the dynamics of gradient-based learning. Our findings reveal that optimal hyper-parameters consistently favour unusually large learning rates. Crucially, we demonstrate that the rapid, aggressive landscape traversal enabled by these large rates not only promotes generalisation-as seen in phenomena like superconvergence-but also statistically minimises the expected computational effort for training. Furthermore, we identify distinct phase transitions in the optimal search strategy: while a single training run suffices for lower accuracy targets, reaching a model's performance limit requires a dramatic shift towards conducting numerous independent, short restarts. Finally, we illustrate how this effort-based paradigm provides a robust framework for model selection, allowing practitioners to choose optimal algorithms based on the difficulty of a problem as perceived by different models for a given target accuracy, or to maximise the achievable accuracy for a fixed budget of gradient descent steps.

Related papers

• Binomial Gradient-Based Meta-Learning for Enhanced Meta-Gradient Estimation [37.795850351021016]
  Gradient-based meta-learning (GBML) relies on gradient descent to adapt the prior to a new task.<n>Existing methods approximate the gradient of prior parameters (meta-gradient) via truncated backpropagation, yet suffer large approximation errors.<n>This work puts forth binomial GBML (BinomGBML), which relies on a truncated binomial expansion for meta-gradient estimation.
  arXiv  Detail & Related papers  (2026-04-14T19:52:15Z)
• Optimizing ML Training with Metagradient Descent [69.89631748402377]
  We introduce an algorithm for efficiently calculating metagradients -- gradients through model training -- at scale.<n>We then introduce a "smooth model training" framework that enables effective optimization using metagradients.
  arXiv  Detail & Related papers  (2025-03-17T22:18:24Z)
• GAQAT: gradient-adaptive quantization-aware training for domain generalization [54.31450550793485]
  We propose a novel Gradient-Adaptive Quantization-Aware Training (GAQAT) framework for DG.<n>Our approach begins by identifying the scale-gradient conflict problem in low-precision quantization.<n>Extensive experiments validate the effectiveness of the proposed GAQAT framework.
  arXiv  Detail & Related papers  (2024-12-07T06:07:21Z)
• LPLgrad: Optimizing Active Learning Through Gradient Norm Sample Selection and Auxiliary Model Training [2.762397703396293]
  Loss Prediction Loss with Gradient Norm (LPLgrad) is designed to quantify model uncertainty effectively and improve the accuracy of image classification tasks. LPLgrad operates in two distinct phases: (i) em Training Phase aims to predict the loss for input features by jointly training a main model and an auxiliary model. This dual-model approach enhances the ability to extract complex input features and learn intrinsic patterns from the data effectively.
  arXiv  Detail & Related papers  (2024-11-20T18:12:59Z)
• What Do Learning Dynamics Reveal About Generalization in LLM Reasoning? [83.83230167222852]
  We find that a model's generalization behavior can be effectively characterized by a training metric we call pre-memorization train accuracy. By connecting a model's learning behavior to its generalization, pre-memorization train accuracy can guide targeted improvements to training strategies.
  arXiv  Detail & Related papers  (2024-11-12T09:52:40Z)
• Gradient-free variational learning with conditional mixture networks [39.827869318925494]
  We introduce CAVI-CMN, a fast, gradient-free variational method for training conditional mixture networks (CMNs)<n>CAVI-CMN achieves competitive and often superior predictive accuracy compared to maximum likelihood estimation (MLE) with backpropagation.<n>As input size or the number of experts increases, computation time scales competitively with MLE.
  arXiv  Detail & Related papers  (2024-08-29T10:43:55Z)
• An Effective Dynamic Gradient Calibration Method for Continual Learning [11.555822066922508]
  Continual learning (CL) is a fundamental topic in machine learning, where the goal is to train a model with continuously incoming data and tasks. Due to the memory limit, we cannot store all the historical data, and therefore confront the catastrophic forgetting'' problem. We develop an effective algorithm to calibrate the gradient in each updating step of the model.
  arXiv  Detail & Related papers  (2024-07-30T16:30:09Z)
• Towards Learning Stochastic Population Models by Gradient Descent [0.0]
  We show that simultaneous estimation of parameters and structure poses major challenges for optimization procedures. We demonstrate accurate estimation of models but find that enforcing the inference of parsimonious, interpretable models drastically increases the difficulty.
  arXiv  Detail & Related papers  (2024-04-10T14:38:58Z)
• Model-Agnostic Multitask Fine-tuning for Few-shot Vision-Language Transfer Learning [59.38343286807997]
  We propose Model-Agnostic Multitask Fine-tuning (MAMF) for vision-language models on unseen tasks. Compared with model-agnostic meta-learning (MAML), MAMF discards the bi-level optimization and uses only first-order gradients. We show that MAMF consistently outperforms the classical fine-tuning method for few-shot transfer learning on five benchmark datasets.
  arXiv  Detail & Related papers  (2022-03-09T17:26:53Z)
• Adapting Stepsizes by Momentumized Gradients Improves Optimization and Generalization [89.66571637204012]
  textscAdaMomentum on vision, and achieves state-the-art results consistently on other tasks including language processing. textscAdaMomentum on vision, and achieves state-the-art results consistently on other tasks including language processing. textscAdaMomentum on vision, and achieves state-the-art results consistently on other tasks including language processing.
  arXiv  Detail & Related papers  (2021-06-22T03:13:23Z)
• 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)

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.