A Lightweight and Gradient-Stable Neural Layer

• URL: http://arxiv.org/abs/2106.04088v4
• Date: Tue, 26 Mar 2024 06:47:13 GMT
• Title: A Lightweight and Gradient-Stable Neural Layer
• Authors: Yueyao Yu, Yin Zhang,
• Abstract summary: We propose a neural-layer architecture based on Householder weighting and absolute-value activating, called Householder-absolute neural layer or simply Han-layer. Compared to a fully connected layer with $d$-neurons and $d$ outputs, a Han-layer reduces the number of parameters and the corresponding computational complexity.
• Score: 3.8263760833282148
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: To enhance resource efficiency and model deployability of neural networks, we propose a neural-layer architecture based on Householder weighting and absolute-value activating, called Householder-absolute neural layer or simply Han-layer. Compared to a fully connected layer with $d$-neurons and $d$ outputs, a Han-layer reduces the number of parameters and the corresponding computational complexity from $O(d^2)$ to $O(d)$. {The Han-layer structure guarantees that the Jacobian of the layer function is always orthogonal, thus ensuring gradient stability (i.e., free of gradient vanishing or exploding issues) for any Han-layer sub-networks.} Extensive numerical experiments show that one can strategically use Han-layers to replace fully connected (FC) layers, reducing the number of model parameters while maintaining or even improving the generalization performance. We will also showcase the capabilities of the Han-layer architecture on a few small stylized models, and discuss its current limitations.

Related papers

• LeanKAN: A Parameter-Lean Kolmogorov-Arnold Network Layer with Improved Memory Efficiency and Convergence Behavior [0.0]
  The Kolmogorov-Arnold network (KAN) is a promising alternative to multi-layer perceptrons (MLPs) for data-driven modeling. Here, we find that MultKAN layers suffer from limited applicability in output layers. We propose LeanKANs, a direct and modular replacement for MultKAN and traditional AddKAN layers.
  arXiv  Detail & Related papers  (2025-02-25T04:43:41Z)
• Graded Neural Networks [0.0]
  This paper presents a novel framework for graded neural networks (GNNs) built over graded vector spaces. Potential applications span machine learning and photonic systems, exemplified by high-speed laser-based implementations.
  arXiv  Detail & Related papers  (2025-02-25T01:08:07Z)
• LESA: Learnable LLM Layer Scaling-Up [57.0510934286449]
  Training Large Language Models (LLMs) from scratch requires immense computational resources, making it prohibitively expensive. Model scaling-up offers a promising solution by leveraging the parameters of smaller models to create larger ones. We propose textbfLESA, a novel learnable method for depth scaling-up.
  arXiv  Detail & Related papers  (2025-02-19T14:58:48Z)
• Kronecker-Factored Approximate Curvature for Modern Neural Network Architectures [85.76673783330334]
  Two different settings of linear weight-sharing layers motivate two flavours of Kronecker-Factored Approximate Curvature (K-FAC) We show they are exact for deep linear networks with weight-sharing in their respective setting. We observe little difference between these two K-FAC variations when using them to train both a graph neural network and a vision transformer.
  arXiv  Detail & Related papers  (2023-11-01T16:37:00Z)
• MuseGNN: Interpretable and Convergent Graph Neural Network Layers at Scale [15.93424606182961]
  We propose a sampling-based energy function and scalable GNN layers that iteratively reduce it, guided by convergence guarantees in certain settings. We also instantiate a full GNN architecture based on these designs, and the model achieves competitive accuracy and scalability when applied to the largest publicly-available node classification benchmark exceeding 1TB in size.
  arXiv  Detail & Related papers  (2023-10-19T04:30:14Z)
• ReLU Neural Networks with Linear Layers are Biased Towards Single- and Multi-Index Models [9.96121040675476]
  This manuscript explores how properties of functions learned by neural networks of depth greater than two layers affect predictions. Our framework considers a family of networks of varying depths that all have the same capacity but different representation costs.
  arXiv  Detail & Related papers  (2023-05-24T22:10:12Z)
• AGNN: Alternating Graph-Regularized Neural Networks to Alleviate Over-Smoothing [29.618952407794776]
  We propose an Alternating Graph-regularized Neural Network (AGNN) composed of Graph Convolutional Layer (GCL) and Graph Embedding Layer (GEL) GEL is derived from the graph-regularized optimization containing Laplacian embedding term, which can alleviate the over-smoothing problem. AGNN is evaluated via a large number of experiments including performance comparison with some multi-layer or multi-order graph neural networks.
  arXiv  Detail & Related papers  (2023-04-14T09:20:03Z)
• WLD-Reg: A Data-dependent Within-layer Diversity Regularizer [98.78384185493624]
  Neural networks are composed of multiple layers arranged in a hierarchical structure jointly trained with a gradient-based optimization. We propose to complement this traditional 'between-layer' feedback with additional 'within-layer' feedback to encourage the diversity of the activations within the same layer. We present an extensive empirical study confirming that the proposed approach enhances the performance of several state-of-the-art neural network models in multiple tasks.
  arXiv  Detail & Related papers  (2023-01-03T20:57:22Z)
• Improved Convergence Guarantees for Shallow Neural Networks [91.3755431537592]
  We prove convergence of depth 2 neural networks, trained via gradient descent, to a global minimum. Our model has the following features: regression with quadratic loss function, fully connected feedforward architecture, RelU activations, Gaussian data instances, adversarial labels. They strongly suggest that, at least in our model, the convergence phenomenon extends well beyond the NTK regime''
  arXiv  Detail & Related papers  (2022-12-05T14:47:52Z)
• BiTAT: Neural Network Binarization with Task-dependent Aggregated Transformation [116.26521375592759]
  Quantization aims to transform high-precision weights and activations of a given neural network into low-precision weights/activations for reduced memory usage and computation. Extreme quantization (1-bit weight/1-bit activations) of compactly-designed backbone architectures results in severe performance degeneration. This paper proposes a novel Quantization-Aware Training (QAT) method that can effectively alleviate performance degeneration.
  arXiv  Detail & Related papers  (2022-07-04T13:25:49Z)
• On Feature Learning in Neural Networks with Global Convergence Guarantees [49.870593940818715]
  We study the optimization of wide neural networks (NNs) via gradient flow (GF) We show that when the input dimension is no less than the size of the training set, the training loss converges to zero at a linear rate under GF. We also show empirically that, unlike in the Neural Tangent Kernel (NTK) regime, our multi-layer model exhibits feature learning and can achieve better generalization performance than its NTK counterpart.
  arXiv  Detail & Related papers  (2022-04-22T15:56:43Z)
• Modeling from Features: a Mean-field Framework for Over-parameterized Deep Neural Networks [54.27962244835622]
  This paper proposes a new mean-field framework for over- parameterized deep neural networks (DNNs) In this framework, a DNN is represented by probability measures and functions over its features in the continuous limit. We illustrate the framework via the standard DNN and the Residual Network (Res-Net) architectures.
  arXiv  Detail & Related papers  (2020-07-03T01:37:16Z)
• High-dimensional Neural Feature Design for Layer-wise Reduction of Training Cost [47.34374677507664]
  We design a ReLU-based multilayer neural network by mapping the feature vectors to a higher dimensional space in every layer. We show that the proposed architecture is norm-preserving and provides an invertible feature vector.
  arXiv  Detail & Related papers  (2020-03-29T15:57:28Z)

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.