Optimal Depth of Neural Networks

• URL: http://arxiv.org/abs/2506.16862v1
• Date: Fri, 20 Jun 2025 09:26:01 GMT
• Title: Optimal Depth of Neural Networks
• Authors: Qian Qi,
• Abstract summary: This paper introduces a formal theoretical framework to address Determining the optimal depth of a neural network.<n>We model the layer-by-layer evolution of hidden representations as a sequential decision process.<n>We propose a novel and practical regularization term, $mathcalL_rm depth$, that encourages the network to learn representations amenable to efficient, early exiting.
• Score: 2.1756081703276
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Determining the optimal depth of a neural network is a fundamental yet challenging problem, typically resolved through resource-intensive experimentation. This paper introduces a formal theoretical framework to address this question by recasting the forward pass of a deep network, specifically a Residual Network (ResNet), as an optimal stopping problem. We model the layer-by-layer evolution of hidden representations as a sequential decision process where, at each layer, a choice is made between halting computation to make a prediction or continuing to a deeper layer for a potentially more refined representation. This formulation captures the intrinsic trade-off between accuracy and computational cost. Our primary theoretical contribution is a proof that, under a plausible condition of diminishing returns on the residual functions, the expected optimal stopping depth is provably finite, even in an infinite-horizon setting. We leverage this insight to propose a novel and practical regularization term, $\mathcal{L}_{\rm depth}$, that encourages the network to learn representations amenable to efficient, early exiting. We demonstrate the generality of our framework by extending it to the Transformer architecture and exploring its connection to continuous-depth models via free-boundary problems. Empirical validation on ImageNet confirms that our regularizer successfully induces the theoretically predicted behavior, leading to significant gains in computational efficiency without compromising, and in some cases improving, final model accuracy.

Related papers

• Neural Collapse is Globally Optimal in Deep Regularized ResNets and Transformers [33.441694038617506]
  We prove that global optima of deep regularized transformers and residual networks (ResNets) with LayerNorm trained with cross entropy or mean squared error loss are approximately collapsed.<n>Our theoretical results are supported by experiments on computer vision and language datasets showing that, as the depth grows, neural collapse indeed becomes more prominent.
  arXiv  Detail & Related papers  (2025-05-21T08:16:03Z)
• Concurrent Training and Layer Pruning of Deep Neural Networks [0.0]
  We propose an algorithm capable of identifying and eliminating irrelevant layers of a neural network during the early stages of training. We employ a structure using residual connections around nonlinear network sections that allow the flow of information through the network once a nonlinear section is pruned.
  arXiv  Detail & Related papers  (2024-06-06T23:19:57Z)
• Fixing the NTK: From Neural Network Linearizations to Exact Convex Programs [63.768739279562105]
  We show that for a particular choice of mask weights that do not depend on the learning targets, this kernel is equivalent to the NTK of the gated ReLU network on the training data. A consequence of this lack of dependence on the targets is that the NTK cannot perform better than the optimal MKL kernel on the training set.
  arXiv  Detail & Related papers  (2023-09-26T17:42:52Z)
• Convergence Guarantees of Overparametrized Wide Deep Inverse Prior [1.5362025549031046]
  Inverse Priors is an unsupervised approach to transform a random input into an object whose image under the forward model matches the observation. We provide overparametrization bounds under which such network trained via continuous-time gradient descent will converge exponentially fast with high probability. This work is thus a first step towards a theoretical understanding of overparametrized DIP networks, and more broadly it participates to the theoretical understanding of neural networks in inverse problem settings.
  arXiv  Detail & Related papers  (2023-03-20T16:49:40Z)
• Globally Optimal Training of Neural Networks with Threshold Activation Functions [63.03759813952481]
  We study weight decay regularized training problems of deep neural networks with threshold activations. We derive a simplified convex optimization formulation when the dataset can be shattered at a certain layer of the network.
  arXiv  Detail & Related papers  (2023-03-06T18:59:13Z)
• Imbedding Deep Neural Networks [0.0]
  Continuous depth neural networks, such as Neural ODEs, have refashioned the understanding of residual neural networks in terms of non-linear vector-valued optimal control problems. We propose a new approach which explicates the network's depth' as a fundamental variable, thus reducing the problem to a system of forward-facing initial value problems.
  arXiv  Detail & Related papers  (2022-01-31T22:00:41Z)
• Path Regularization: A Convexity and Sparsity Inducing Regularization for Parallel ReLU Networks [75.33431791218302]
  We study the training problem of deep neural networks and introduce an analytic approach to unveil hidden convexity in the optimization landscape. We consider a deep parallel ReLU network architecture, which also includes standard deep networks and ResNets as its special cases.
  arXiv  Detail & Related papers  (2021-10-18T18:00:36Z)
• Layer Adaptive Node Selection in Bayesian Neural Networks: Statistical Guarantees and Implementation Details [0.5156484100374059]
  Sparse deep neural networks have proven to be efficient for predictive model building in large-scale studies. We propose a Bayesian sparse solution using spike-and-slab Gaussian priors to allow for node selection during training. We establish the fundamental result of variational posterior consistency together with the characterization of prior parameters.
  arXiv  Detail & Related papers  (2021-08-25T00:48:07Z)
• Non-Gradient Manifold Neural Network [79.44066256794187]
  Deep neural network (DNN) generally takes thousands of iterations to optimize via gradient descent. We propose a novel manifold neural network based on non-gradient optimization.
  arXiv  Detail & Related papers  (2021-06-15T06:39:13Z)
• A Convergence Theory Towards Practical Over-parameterized Deep Neural Networks [56.084798078072396]
  We take a step towards closing the gap between theory and practice by significantly improving the known theoretical bounds on both the network width and the convergence time. We show that convergence to a global minimum is guaranteed for networks with quadratic widths in the sample size and linear in their depth at a time logarithmic in both. Our analysis and convergence bounds are derived via the construction of a surrogate network with fixed activation patterns that can be transformed at any time to an equivalent ReLU network of a reasonable size.
  arXiv  Detail & Related papers  (2021-01-12T00:40:45Z)
• Generalization bound of globally optimal non-convex neural network training: Transportation map estimation by infinite dimensional Langevin dynamics [50.83356836818667]
  We introduce a new theoretical framework to analyze deep learning optimization with connection to its generalization error. Existing frameworks such as mean field theory and neural tangent kernel theory for neural network optimization analysis typically require taking limit of infinite width of the network to show its global convergence.
  arXiv  Detail & Related papers  (2020-07-11T18:19:50Z)

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.