Convolutions and More as Einsum: A Tensor Network Perspective with Advances for Second-Order Methods

• URL: http://arxiv.org/abs/2307.02275v2
• Date: Wed, 23 Oct 2024 22:47:01 GMT
• Title: Convolutions and More as Einsum: A Tensor Network Perspective with Advances for Second-Order Methods
• Authors: Felix Dangel,
• Abstract summary: We simplify convolutions by viewing them as tensor networks (TNs) TNs allow reasoning about the underlying tensor multiplications by drawing diagrams, manipulating them to perform function transformations like differentiation, and efficiently evaluating them with einsum. Our TN implementation accelerates KFAC variant up to 4.5x while removing the standard implementation's memory overhead, and enables new hardware-efficient dropouts for approximate backpropagation.
• Score: 2.8645507575980074
• License:
• Abstract: Despite their simple intuition, convolutions are more tedious to analyze than dense layers, which complicates the transfer of theoretical and algorithmic ideas to convolutions. We simplify convolutions by viewing them as tensor networks (TNs) that allow reasoning about the underlying tensor multiplications by drawing diagrams, manipulating them to perform function transformations like differentiation, and efficiently evaluating them with einsum. To demonstrate their simplicity and expressiveness, we derive diagrams of various autodiff operations and popular curvature approximations with full hyper-parameter support, batching, channel groups, and generalization to any convolution dimension. Further, we provide convolution-specific transformations based on the connectivity pattern which allow to simplify diagrams before evaluation. Finally, we probe performance. Our TN implementation accelerates a recently-proposed KFAC variant up to 4.5x while removing the standard implementation's memory overhead, and enables new hardware-efficient tensor dropout for approximate backpropagation.

Related papers

• Variable-size Symmetry-based Graph Fourier Transforms for image compression [65.7352685872625]
  We propose a new family of Symmetry-based Graph Fourier Transforms of variable sizes into a coding framework. Our proposed algorithm generates symmetric graphs on the grid by adding specific symmetrical connections between nodes. Experiments show that SBGFTs outperform the primary transforms integrated in the explicit Multiple Transform Selection.
  arXiv  Detail & Related papers  (2024-11-24T13:00:44Z)
• conv_einsum: A Framework for Representation and Fast Evaluation of Multilinear Operations in Convolutional Tensorial Neural Networks [28.416123889998243]
  We develop a framework for representing tensorial convolution layers as einsum-like strings and a meta-algorithm conv_einsum which is able to evaluate these strings in a FLOPs-minimizing manner.
  arXiv  Detail & Related papers  (2024-01-07T04:30:12Z)
• Deep Neural Networks with Efficient Guaranteed Invariances [77.99182201815763]
  We address the problem of improving the performance and in particular the sample complexity of deep neural networks. Group-equivariant convolutions are a popular approach to obtain equivariant representations. We propose a multi-stream architecture, where each stream is invariant to a different transformation.
  arXiv  Detail & Related papers  (2023-03-02T20:44:45Z)
• Empowering Networks With Scale and Rotation Equivariance Using A Similarity Convolution [16.853711292804476]
  We devise a method that endows CNNs with simultaneous equivariance with respect to translation, rotation, and scaling. Our approach defines a convolution-like operation and ensures equivariance based on our proposed scalable Fourier-Argand representation. We validate the efficacy of our approach in the image classification task, demonstrating its robustness and the generalization ability to both scaled and rotated inputs.
  arXiv  Detail & Related papers  (2023-03-01T08:43:05Z)
• OneDConv: Generalized Convolution For Transform-Invariant Representation [76.15687106423859]
  We propose a novel generalized one dimension convolutional operator (OneDConv) It dynamically transforms the convolution kernels based on the input features in a computationally and parametrically efficient manner. It improves the robustness and generalization of convolution without sacrificing the performance on common images.
  arXiv  Detail & Related papers  (2022-01-15T07:44:44Z)
• Revisiting Transformation Invariant Geometric Deep Learning: Are Initial Representations All You Need? [80.86819657126041]
  We show that transformation-invariant and distance-preserving initial representations are sufficient to achieve transformation invariance. Specifically, we realize transformation-invariant and distance-preserving initial point representations by modifying multi-dimensional scaling. We prove that TinvNN can strictly guarantee transformation invariance, being general and flexible enough to be combined with the existing neural networks.
  arXiv  Detail & Related papers  (2021-12-23T03:52:33Z)
• Self-Supervised Graph Representation Learning via Topology Transformations [61.870882736758624]
  We present the Topology Transformation Equivariant Representation learning, a general paradigm of self-supervised learning for node representations of graph data. In experiments, we apply the proposed model to the downstream node and graph classification tasks, and results show that the proposed method outperforms the state-of-the-art unsupervised approaches.
  arXiv  Detail & Related papers  (2021-05-25T06:11:03Z)
• Adaptive Learning of Tensor Network Structures [6.407946291544721]
  We leverage the TN formalism to develop a generic and efficient adaptive algorithm to learn the structure and the parameters of a TN from data. Our algorithm can adaptively identify TN structures with small number of parameters that effectively optimize any differentiable objective function.
  arXiv  Detail & Related papers  (2020-08-12T16:41:56Z)
• Supervised Learning for Non-Sequential Data: A Canonical Polyadic Decomposition Approach [85.12934750565971]
  Efficient modelling of feature interactions underpins supervised learning for non-sequential tasks. To alleviate this issue, it has been proposed to implicitly represent the model parameters as a tensor. For enhanced expressiveness, we generalize the framework to allow feature mapping to arbitrarily high-dimensional feature vectors.
  arXiv  Detail & Related papers  (2020-01-27T22:38:40Z)

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.