A new tensor-based layer reduces neural network dimensions without losing important features.
problem Reducing dimensionality in tensor-structured feature data for deep neural networks.
method TensorProjection layer that projects input tensors into output tensors with reduced dimensions through mode-wise projections.
result The TensorProjection layer outperforms traditional downsampling methods in tasks like medical image classification and segmentation.
Tensor regression networks achieve high compression rate of neural networks while having slight impact on performances. They do so by imposing low tensor rank structure on the weight matrices of fully connected layers. In recent years, tensor regression networks have been investigated from the perspective of their comp…
TWIST algorithm detects communities in multi-layer networks with tensor decomposition.
problem Community detection in multi-layer networks with multiple node-modality relationships.
method Tensor-based TWIST algorithm for global/local node and layer memberships.
result Accurate community detection with small misclassification error as network size increases.
New method compresses deep learning layers using tensor decomposition.
problem Reduction of computation cost and interpretability for tensor data.
method CP-decomposition to compress convolutional layers in deep learning.
result Reduces model complexity and maintains prediction performance.
A new tensor network method for image classification reduces computation cost.
problem Efficiently classifying images in high-dimensional spaces.
method Proposes a multi-layered tensor network (MLTN) that performs one MPS operation per layer, reducing computation cost.
result Reduces computation cost without degrading performance.
Paper analyzes infinite-width attention layers using Tensor Programs.
problem Capturing the infinite-width limit of attention layers.
method Tensor Programs framework to rigorously identify the limit distribution.
result Derives exact form of infinite-width limit distribution without Gaussian approximations.
This work introduces a tensor-based method to perform supervised classification on spatiotemporal data processed in an echo state network. Typically when performing supervised classification tasks on data processed in an echo state network, the entire collection of hidden layer node states from the training dataset is …
Characterizes a specific type of neural network for alternating group equivariance.
problem Understanding and characterizing neural networks with alternating group equivariance.
method Characterization of all possible An-equivariant neural networks using tensor powers of Rn. result Found a basis of matrices for learnable, linear An-equivariant layer functions. Recurrent Neural Networks (RNNs), which are a powerful scheme for modeling temporal and sequential data need to capture long-term dependencies on datasets and represent them in hidden layers with a powerful model to capture more information from inputs. For modeling long-term dependencies in a dataset, the gating mecha…
Tensor dropout improves CNN robustness and efficiency.
problem Limited generalization and robustness of deep CNNs.
method Tensor layers with low-rank structures and tensor dropout.
result Tensor dropout outperforms other methods in robustness and efficiency.
Tensor network architecture for classification and regression using wavelet transformations.
problem Efficiently performing classification and regression tasks on complex data.
method Tensor network layers based on MERA and MPS, with adaptive fine-graining.
result Adaptive fine-graining improves model performance without loss in accuracy.
Tensor models decode human perception and memory using SPO triples.
problem Understanding implicit and explicit perception and memory in the brain.
method Tensor models with SPO triples, dual representations, and four layers.
result Semantic memory is crucial for explicit perception and declarative memories.
The paper analyzes tensor recovery from symmetric rank-one measurements using information theory.
problem Recovering tensors with low symmetric rank from symmetric rank-one measurements.
method Covering numbers argument, Carbery-Wright inequality, orthogonal polynomials, Fano's inequality.
result Near-optimal sample complexity bounds for log-concave distributions.
Characterizes group-equivariant neural networks for three groups.
problem Understanding equivariant neural networks for orthogonal, special orthogonal, and symplectic groups.
method Characterized all possible group-equivariant neural networks for three groups.
result Found spanning sets of matrices for learnable, linear equivariant layer functions.
Generative model improves tabular data density estimation.
problem Challenges in estimating tabular data distribution.
method Tensor contraction layers and transformers in VAEs.
result Embedding representations improve density estimation metrics.
We introduce a new parameterization method for deep learning layers using spectral tensor train decomposition.
problem Efficiency and stability in deep learning models with weight matrix compression.
method Spectral Tensor Train Parameterization (STTP) of weight matrices.
result Improved compression and training stability in neural networks.
Paper introduces TSSDMN for modeling dynamic multilayer networks.
problem Capturing temporal and cross-layer dynamics in multilayer networks.
method Tensor State Space Model (TSSDMN) using symmetric Tucker decomposition.
result TSSDMN uniquely captures temporal dynamics within and across layers.
Recently, deep neural networks (DNNs) have been regarded as the state-of-the-art classification methods in a wide range of applications, especially in image classification. Despite the success, the huge number of parameters blocks its deployment to situations with light computing resources. Researchers resort to the re…
New IP analysis for deep neural networks using Rényi's entropy and tensor kernels.
problem Estimating mutual information in high-dimensional hidden layers of deep neural networks.
method Matrix-based Rényi's entropy coupled with tensor kernels for convolutional layers.
result First comprehensive IP analysis of large-scale DNNs and CNNs.
Long Short-Term Memory (LSTM) is a popular approach to boosting the ability of Recurrent Neural Networks to store longer term temporal information. The capacity of an LSTM network can be increased by widening and adding layers. However, usually the former introduces additional parameters, while the latter increases the…
A method predicts GNS of transformer layers using normalization layer norms.
problem Estimating gradient noise scale with minimal variance.
method Simultaneously compute per-example gradient norms and parameter gradients.
result Total GNS is predicted well by normalization layer GNS.
Inspired by coarse-graining approaches used in physics, we show how similar algorithms can be adapted for data. The resulting algorithms are based on layered tree tensor networks and scale linearly with both the dimension of the input and the training set size. Computing most of the layers with an unsupervised algorith…
New basis for permutation equivariant layers reduces computation costs.
problem Efficiently computing permutation equivariant layers in neural networks.
method Generalized partition algebra basis with low-rank tensors.
result Low-rank tensors enable faster computation compared to orbit basis.
Layered graphical models improve discriminative learning efficiency.
problem Improving discriminative learning efficiency in graphical models.
method Designing layered graphical models (LGMs) in analogy to neural networks, using tensorized truncated variational inference and backpropagation.
result LGMs achieve competitive results in image classification, comparable to neural networks.
Automatically designs normalization and activation layers together.
problem Designing normalization and activation layers separately.
method Unified tensor-to-tensor computation graph, low-level mathematical functions, rejection protocols, multi-objective evolution.
result Discovery of EvoNorms with novel structures.
Deep neural networks have demonstrated state-of-the-art performance in a variety of real-world applications. In order to obtain performance gains, these networks have grown larger and deeper, containing millions or even billions of parameters and over a thousand layers. The trade-off is that these large architectures r…
We establish connections between the problem of learning a two-layer neural network and tensor decomposition. We consider a model with feature vectors x∈Rd, r hidden units with weights {wi}1≤i≤r and output y∈R, i.e., $y=\sum_{i=1}^r σ( \boldsymbol w_i…
Optimizes neural network training by dynamically updating Tucker decomposition ranks.
problem Redundant parameters in neural network architectures.
method Geometry-aware training of factorized layers in tensor Tucker format.
result Optimal locally approximating the original dynamics without initial rank knowledge.
Paper perfect clusters sparse, diverse multilayer networks.
problem Clustering sparse, diverse multilayer networks.
method Tensor-based methodology pooling all layers' information.
result Achieves perfect clustering under sparser conditions than previous models.
New tensor network decompositions improve CNN performance.
problem Limited exploration of tensor network decompositions for CNNs.
method Characterized a new class of CNN modules and experimentally compared various decompositions.
result Some nonlinear decompositions outperform existing ones in terms of accuracy and efficiency.
We identify and approximate weights of two-layer neural networks from few samples.
problem Identifying and approximating weights of two-layer neural networks from limited data.
method Active sampling of finite difference approximations to Hessians, solving robust nonlinear programs, and gradient descent.
result Stable recovery of network weights under verifiable conditions.
New framework finds more efficient linear layers over structured matrices.
problem Efficient alternatives for dense linear layers in neural networks.
method Unified framework searching over all linear operators, developing a taxonomy based on computational and algebraic properties.
result BTT-MoE provides substantial compute-efficiency gains over dense layers and standard MoE.
Algorithm learns polynomial transformations of Gaussian distributions.
problem Learning high-dimensional polynomial transformations of Gaussian distributions.
method Polynomial-time algorithms for smoothed settings, tensor ring decomposition.
result First end-to-end guarantees for learning pushforwards under neural networks.
Local invertibility of ray transforms on convex manifolds.
problem Invertibility of ray transforms on compact Riemannian manifolds with strictly convex boundary.
method Local invertibility results for transverse and mixed ray transforms of 1 and 1+1 tensors.
result Local invertibility of ray transforms near boundary points, leading to global results.
Alternative to convolutions using decision trees for neural networks.
problem Replacing complex convolutions with simpler decision-based layers.
method Binary decisions as indices to conditional distributions, trained using backpropagation.
result Performance similar to conventional neural networks, with runtime improvements.
Category theory enhances understanding of group-equivariant neural networks.
problem Understanding and working with group-equivariant neural networks.
method Application of category theory to tensor power spaces of Rn for groups Sn, O(n), Sp(n), and SO(n). result New insights and an algorithm for computing equivariant linear layers.
Paper compresses ViTs models by 60% with minimal accuracy loss.
problem Memory constraints in ViTs models.
method Activation-aware low-rank tensor approximations.
result 60% reduction in model size with <1% accuracy loss.
Training neural networks is a challenging non-convex optimization problem, and backpropagation or gradient descent can get stuck in spurious local optima. We propose a novel algorithm based on tensor decomposition for guaranteed training of two-layer neural networks. We provide risk bounds for our proposed method, with…
Paper optimizes clustering for multi-layer networks and discrete mixtures.
problem Optimizing clustering in multi-layer networks and discrete mixtures.
method Two-stage method: tensor-based initialization and likelihood-based refinement.
result Achieves minimax optimal error rate for multi-layer networks and discrete mixtures.
A new algorithm reduces memory usage for deep learning models.
problem Training deep learning models requires significant memory.
method Dynamic Tensor Rematerialization (DTR) is a greedy online algorithm that dynamically plans recomputations.
result DTR achieves comparable performance to optimal static checkpointing with only a small memory budget.
Tensor Train layer improves BLEU scores in NMT models.
problem Improving Neural Machine Translation (NMT) models' performance.
method Implemented Tensor Train layer in TensorFlow for NMT training.
result Higher learning rates and more 'rectangular' core dimensions improve BLEU scores.
Study on identifiability of deep polynomial neural networks.
problem Understanding when polynomial neural networks can be uniquely identified.
method Comprehensive analysis including various architectures, using tensor decompositions and Kruskal-type theorems.
result Identifiability conditions for deep PNNs, including layer width and activation degree constraints.
ALMA improves clustering of multilayer networks.
problem Clustering multilayer networks with distinct layers and communities.
method Alternating minimization algorithm (ALMA) for simultaneous layer partition and community estimation.
result ALMA achieves higher accuracy than TWIST in clustering multilayer networks.
Tensor neural network improves human pose classification from 3D skeleton data.
problem Efficiently processing spatiotemporal data for human pose classification.
method Proposes a tensor-based neural network with three components: spatiotemporal feature construction, tensor fusion, and tensor-based neural network processing.
result Achieves state-of-the-art performance in human pose classification.
It is well known that multiplication operations in convolutional layers of common CNNs consume a lot of time during inference stage. In this article we present a flexible method to decrease both computational complexity of convolutional layers in inference as well as amount of space to store them. The method is based o…
New method constructs multilayer networks from financial data, capturing dependencies across different risk factors.
problem Difficult construction of multilayer networks, neglecting time delays and interdependencies.
method Tucker tensor autoregression for direct multilayer network construction.
result Captures within and between connections, identifies strong interconnections between volumes and prices layers.
A new sketching method reduces tensor memory usage and enables efficient tensor operations.
problem Efficiently compressing and retaining tensor structure in large datasets.
method Higher-order Count Sketch (HCS) using multiple hash functions and tensor products.
result HCS achieves significant memory savings and efficient tensor operations.
Polynomial fusion layer improves speech-driven facial animation.
problem Recent facial synthesis relies on low-dimensional representations and concatenation, ignoring higher-order interactions.
method Proposes a polynomial fusion layer to model higher-order interactions of facial encodings.
result Demonstrates improved video quality, audiovisual synchronisation, and blink generation.