Abide by the Law and Follow the Flow: Conservation Laws for Gradient Flows

• URL: http://arxiv.org/abs/2307.00144v2
• Date: Wed, 10 Jul 2024 13:15:10 GMT
• Title: Abide by the Law and Follow the Flow: Conservation Laws for Gradient Flows
• Abstract summary: "Conservation laws" define quantities conserved during gradient flows of a given model. "implicit bias" is believed to be responsible for some favorable properties of trained models. We show how to find the maximal number of independent conservation laws.
• Score: 28.287184613608435
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Understanding the geometric properties of gradient descent dynamics is a key ingredient in deciphering the recent success of very large machine learning models. A striking observation is that trained over-parameterized models retain some properties of the optimization initialization. This "implicit bias" is believed to be responsible for some favorable properties of the trained models and could explain their good generalization properties. The purpose of this article is threefold. First, we rigorously expose the definition and basic properties of "conservation laws", that define quantities conserved during gradient flows of a given model (e.g. of a ReLU network with a given architecture) with any training data and any loss. Then we explain how to find the maximal number of independent conservation laws by performing finite-dimensional algebraic manipulations on the Lie algebra generated by the Jacobian of the model. Finally, we provide algorithms to: a) compute a family of polynomial laws; b) compute the maximal number of (not necessarily polynomial) independent conservation laws. We provide showcase examples that we fully work out theoretically. Besides, applying the two algorithms confirms for a number of ReLU network architectures that all known laws are recovered by the algorithm, and that there are no other independent laws. Such computational tools pave the way to understanding desirable properties of optimization initialization in large machine learning models.

Related papers

• SGD in Multiclass Logistic Regression: Sequential Learning and Scaling Laws [55.81298869922287]
  We study the training dynamics of multiclass logistic regression on high-dimensional Gaussian mixture models with a large number of classes.<n>We establish precise scaling laws governing the cross-entropy risk under gradient-based optimization.
  arXiv  Detail & Related papers  (2026-09-07T18:30:09Z)
• Learning to Discover Iterative Spectral Algorithms [46.39331984989963]
  AutoSpec is a neural network framework for discovering iterative spectral algorithms for large-scale numerical linear algebra.<n>We apply AutoSpec to discovering algorithms for representative numerical linear algebra tasks.<n>On real-world solvers, the learned procedures deliver orders-of-magnitude improvements in accuracy and/or reductions in count.
  arXiv  Detail & Related papers  (2026-02-10T08:39:59Z)
• Towards Robust Scaling Laws for Optimizers [89.21160945066737]
  Empirical scaling laws are widely used to predict loss as model size and training data grow.<n>We show that Chinchilla-style scaling laws emerge naturally as a result of loss decomposition into irreducible, approximation, and optimization errors.
  arXiv  Detail & Related papers  (2026-02-07T21:40:33Z)
• Beyond Memorization: Extending Reasoning Depth with Recurrence, Memory and Test-Time Compute Scaling [60.63703438729223]
  We show how different architectures and training methods affect model multi-step reasoning capabilities.<n>We confirm that increasing model depth plays a crucial role for sequential computations.
  arXiv  Detail & Related papers  (2025-08-22T18:57:08Z)
• Efficient Knowledge Graph Unlearning with Zeroth-order Information [16.462136830216856]
  We present an efficient knowledge graph (KG) unlearning algorithm.<n> KG unlearning is nontrivial due to the distinctive structure of KG and the semantic relations between entities.<n>Our experimental results demonstrate that the proposed method outperforms other state-of-the-art graph unlearning baselines.
  arXiv  Detail & Related papers  (2025-08-19T17:22:50Z)
• Algebraic Machine Learning: Learning as computing an algebraic decomposition of a task [41.94295877935867]
  We propose an alternative foundation based on Abstract Algebra, with mathematics that facilitates the analysis of learning. In this approach, the goal of the task and the data are encoded as axioms of an algebra, and a model is obtained where only these axioms and their logical consequences hold. We validate this new learning principle on standard datasets such as MNIST, FashionMNIST, CIFAR-10, and medical images, achieving performance comparable to optimized multilayer perceptrons.
  arXiv  Detail & Related papers  (2025-02-27T10:13:42Z)
• Scaling Law for Stochastic Gradient Descent in Quadratically Parameterized Linear Regression [5.801904710149222]
  In machine learning, the scaling law describes how the model performance improves with the model and data size scaling up. This paper studies the scaling law over a linear regression with the model being quadratically parameterized. As a result, in the canonical linear regression, we provide explicit separations for curves between generalization with and without feature learning, and the information-theoretical lower bound that is to parametrization method and the algorithm.
  arXiv  Detail & Related papers  (2025-02-13T09:29:04Z)
• Random Sparse Lifts: Construction, Analysis and Convergence of finite sparse networks [17.487761710665968]
  We present a framework to define a large class of neural networks for which, by construction, training by gradient flow provably reaches arbitrarily low loss when the number of parameters grows.
  arXiv  Detail & Related papers  (2025-01-10T12:52:00Z)
• Gradient Estimation and Variance Reduction in Stochastic and Deterministic Models [0.0]
  This dissertation considers unconstrained, nonlinear optimization problems. We focus on the gradient itself, that key quantity which enables the solution of such problems. We present a new framework for calculating the gradient of problems involving both deterministic and elements.
  arXiv  Detail & Related papers  (2024-05-14T14:41:58Z)
• Scaling and renormalization in high-dimensional regression [72.59731158970894]
  This paper presents a succinct derivation of the training and generalization performance of a variety of high-dimensional ridge regression models. We provide an introduction and review of recent results on these topics, aimed at readers with backgrounds in physics and deep learning.
  arXiv  Detail & Related papers  (2024-05-01T15:59:00Z)
• Efficient and Generalizable Certified Unlearning: A Hessian-free Recollection Approach [8.875278412741695]
  Machine unlearning strives to uphold the data owners' right to be forgotten by enabling models to selectively forget specific data. We develop an algorithm that achieves near-instantaneous unlearning as it only requires a vector addition operation.
  arXiv  Detail & Related papers  (2024-04-02T07:54:18Z)
• Towards Optimal Learning of Language Models [124.65669486710992]
  We present a theory for the optimal learning of language models (LMs) We derive a theorem, named Learning Law, to reveal the properties of the dynamics in the optimal learning process under our objective. We empirically verify that the optimal learning of LMs essentially stems from the improvement of the coefficients in the scaling law of LMs.
  arXiv  Detail & Related papers  (2024-02-27T18:52:19Z)
• Covering Number of Real Algebraic Varieties and Beyond: Improved Bounds and Applications [8.438718130535296]
  We prove upper bounds on the covering number of sets in Euclidean space. We show that bounds improve the best known general bound by Yomdin-Comte. We illustrate the power of the result on three computational applications.
  arXiv  Detail & Related papers  (2023-11-09T03:06:59Z)
• CoLA: Exploiting Compositional Structure for Automatic and Efficient Numerical Linear Algebra [62.37017125812101]
  We propose a simple but general framework for large-scale linear algebra problems in machine learning, named CoLA. By combining a linear operator abstraction with compositional dispatch rules, CoLA automatically constructs memory and runtime efficient numerical algorithms. We showcase its efficacy across a broad range of applications, including partial differential equations, Gaussian processes, equivariant model construction, and unsupervised learning.
  arXiv  Detail & Related papers  (2023-09-06T14:59:38Z)
• Graph Polynomial Convolution Models for Node Classification of Non-Homophilous Graphs [52.52570805621925]
  We investigate efficient learning from higher-order graph convolution and learning directly from adjacency matrix for node classification. We show that the resulting model lead to new graphs and residual scaling parameter. We demonstrate that the proposed methods obtain improved accuracy for node-classification of non-homophilous parameters.
  arXiv  Detail & Related papers  (2022-09-12T04:46:55Z)
• A framework for overparameterized learning [0.0]
  An explanation for the success of deep neural networks is a central question in theoretical machine learning. We propose a framework consisting of a prototype learning problem, which is general enough to cover many popular problems. We then demonstrate that supervised learning, variational autoencoders and training with gradient penalty can be translated to the prototype problem.
  arXiv  Detail & Related papers  (2022-05-26T17:17:46Z)
• Fractal Structure and Generalization Properties of Stochastic Optimization Algorithms [71.62575565990502]
  We prove that the generalization error of an optimization algorithm can be bounded on the complexity' of the fractal structure that underlies its generalization measure. We further specialize our results to specific problems (e.g., linear/logistic regression, one hidden/layered neural networks) and algorithms.
  arXiv  Detail & Related papers  (2021-06-09T08:05:36Z)
• Towards Lower Bounds on the Depth of ReLU Neural Networks [7.355977594790584]
  We investigate whether the class of exactly representable functions strictly increases by adding more layers. We settle an old conjecture about piecewise linear functions by Wang and Sun (2005) in the affirmative. We present upper bounds on the sizes of neural networks required to represent functions with logarithmic depth.
  arXiv  Detail & Related papers  (2021-05-31T09:49:14Z)
• Reinforcement Learning with General Value Function Approximation: Provably Efficient Approach via Bounded Eluder Dimension [124.7752517531109]
  We establish a provably efficient reinforcement learning algorithm with general value function approximation. We show that our algorithm achieves a regret bound of $widetildeO(mathrmpoly(dH)sqrtT)$ where $d$ is a complexity measure. Our theory generalizes recent progress on RL with linear value function approximation and does not make explicit assumptions on the model of the environment.
  arXiv  Detail & Related papers  (2020-05-21T17:36:09Z)

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.