Paper investigates Lipschitz constants of self-attention modules in neural networks.
problem Lipschitz constants of self-attention modules in neural networks.
method Proved standard dot-product self-attention is not Lipschitz for unbounded input domain. Proposed L2 self-attention that is Lipschitz. Derived upper bound on L2 self-attention's Lipschitz constant.
result Proved standard self-attention is not Lipschitz for unbounded input domain and proposed an alternative L2 self-attention that is Lipschitz.
A new neural network separates vocals from music accompaniment.
problem Separating vocals from music accompaniment in recordings.
method Self-attention convolutional neural network (CNN) with densely-connected blocks.
result 19.5% relative improvement in vocals separation.
TSAM predicts directed temporal links using GCN and self-attention.
problem Predicting links in directed temporal networks.
method GCN, self-attention mechanism, autoencoder architecture, graph attentional layers, graph convolutional layers, graph recurrent unit layer.
result TSAM outperforms benchmarks on four realistic networks.
Derives an empirical capacity model for self-attention neural networks.
problem Theoretical capacity of large transformer models is not fully utilized by current optimization algorithms.
method Analyzes memory capacity of transformers using synthetic training data and common training algorithms.
result Derives an empirical capacity model (ECM) for a generic transformer.
Transformers' self-attention mechanism is mapped to a generalized Potts model.
problem Uncertainty in what type of data distribution self-attention can efficiently learn.
method Decouple word positions and embeddings, then show self-attention learns a generalized Potts model.
result Training self-attention is equivalent to solving the inverse Potts problem.
Self-attention improves satellite time series classification without preprocessing.
problem Efficiently classifying raw satellite time series data.
method Comparison of deep learning models including self-attention, 1D-convolutions, recurrence, and random forest.
result Self-attention and recurrent neural networks outperform convolutional neural networks on raw satellite time series.
LieTransformer extends self-attention to Lie groups for improved deep learning tasks.
problem Improving deep learning performance through group equivariant self-attention.
method LieSelfAttention layers that are equivariant to arbitrary Lie groups and their discrete subgroups.
result Competitive experimental results on various tasks.
Study reveals self-attention's role in learning and generalizing interactions.
problem Understanding self-attention's theoretical role in neural architectures.
method Interacting entities analysis, including multi-agent RL and genetic sequences.
result Self-attention efficiently represents, learns, and generalizes pairwise interactions.
Aligns attention distributions for improved accuracy and robustness.
problem Improving the accuracy and robustness of neural networks using attention mechanisms.
method Alignment attention that encourages key and query distributions to match within each head.
result Alignment attention leads to better accuracy, uncertainty estimation, and robustness across various tasks.
Insurance companies must manage millions of claims per year. While most of these claims are non-fraudulent, fraud detection is core for insurance companies. The ultimate goal is a predictive model to single out the fraudulent claims and pay out the non-fraudulent ones immediately. Modern machine learning methods are we…
Mathematical framework for understanding attention in neural networks.
problem Lack of theoretical understanding of attention in neural networks.
method Proposes a measure-theoretic model of attention and interprets self-attention as a system of self-interacting particles.
result Shows that attention is Lipschitz-continuous under suitable assumptions.
This paper offers an overview of neural network compression techniques.
problem Overparameterized neural networks are large and resource-intensive.
method Pruning, quantization, tensor decomposition, knowledge distillation.
result A comprehensive review of compression techniques for deep neural networks.
Efficient neural network optimization reduces costs and improves model performance.
problem High computational costs in optimizing neural networks, especially at scale.
method Introduces self-attentive feed-forward neural units (SAFFU) for efficient optimization.
result Explicit solutions outperform models optimized by backpropagation alone, and further training with backpropagation leads to better optima from smaller data sets.
Analyzes self-attention in recurrent networks, proving it mitigates vanishing gradients.
problem Vanishing gradients in recurrent networks when capturing long-term dependencies.
method Formal analysis of self-attention's effect on gradient propagation, proposing a relevancy screening mechanism.
result Self-attention mitigates vanishing gradients in recurrent networks, providing guarantees.
Improved electrical load forecasting model using Fourier-enhanced RNN.
problem Electrical load time series downscaling with high accuracy and low error.
method Combines recurrent neural network with Fourier seasonal embeddings and self-attention.
result Significantly reduces RMSE across different time horizons compared to existing methods.
The paper investigates how Global Self-attention improves GCNs.
problem Improving expressive power and addressing overfitting in GCNs.
method Applying Global Self-attention mechanism over graph features.
result GSA mechanism enhances expressive power and mitigates overfitting and over-smoothing in GCNs.
Hyper-SAGNN learns patterns in hypergraphs for complex interactions.
problem Learning patterns in hypergraphs with variable-sized heterogeneous hyperedges.
method Self-attention based graph neural network for homogeneous and heterogeneous hypergraphs.
result Significantly outperforms state-of-the-art methods on various tasks.
Graph neural networks speed up nonnegative matrix factorization.
problem Efficiently factorize nonnegative matrices for various applications.
method Developed a graph neural network that combines bipartite self-attention with ADMM updates.
result Significant acceleration achieved in nonnegative matrix factorization.
A framework for transformer attention layers derived from SVR.
problem Developing principled attention mechanisms for transformers.
method Mapping self-attention to SVR, deriving new attention types.
result Improved transformer performance and efficiency.
Quaternion self-attention reduces computational cost and improves performance.
problem Existing quaternion self-attention increases computational cost and diverges attention distributions.
method Proposes a shared-score quaternion self-attention mechanism.
result Reduces score-computation multiplications by 75% and softmax operations from four to one.
Self-attention model improves HAR from wearable sensors.
problem Capturing spatio-temporal context from sensor data.
method Proposes a self-attention based neural network model.
result Significant performance improvement over state-of-the-art models.
Learning latent representations of nodes in graphs is an important and ubiquitous task with widespread applications such as link prediction, node classification, and graph visualization. Previous methods on graph representation learning mainly focus on static graphs, however, many real-world graphs are dynamic and evol…
Self-attention models benefit equally from width and depth, but beyond a certain point, depth becomes less efficient.
problem Understanding the optimal balance between depth and width in self-attention models.
method Theoretical predictions and empirical ablations on networks of varying depths and widths.
result An optimal width of 30K is recommended for a 1-Trillion parameter network, marking a significant width for self-attention models.
GSA-Nets apply group equivariance to self-attention for vision tasks.
problem Improving self-attention networks for vision tasks.
method Define group-equivariant positional encodings.
result GSA-Nets outperform non-equivariant self-attention networks on vision benchmarks.
AttViz offers online visualizations of neural language model attention mechanisms.
problem Limited interpretability of neural language models.
method Online toolkit for exploring self-attention mechanisms.
result Visualizations help understand model decision-making.
New model mimics neural next item recommendation using Hankel matrices.
problem Next item recommendation efficiency and structural knowledge capture.
method Tensor factorization with Hankel matrix representation.
result Model performs competitively with neural networks but is simpler.
Develops first robustness verification for complex Transformers.
problem Certify prediction behavior of Transformers with complex self-attention layers.
method Resolves challenges of cross-nonlinearity and cross-position dependency in Transformers.
result Certified robustness bounds are significantly tighter than Interval Bound Propagation.
Proposes a novel SAM operator for separate item and relational memories.
problem Limited memory interactions in neural networks.
method Introduces a Self-attentive Associative Memory (SAM) operator to separate item and relational memories.
result Achieves competitive results in various tasks, including geometry, graph, reinforcement learning, and question answering.
Self-attention prefers sparse functions of input sequences, reducing sample complexity.
problem Understanding the inductive biases of self-attention in modeling long-range dependencies.
method Theoretical analysis and synthetic experiments to probe sample complexity of learning sparse functions with Transformers.
result Bounded-norm Transformer networks can represent sparse functions of the input sequence with logarithmic sample complexity.
SANST uses self-attentive networks with spatial and temporal embeddings for better POI recommendations.
problem Next point-of-interest (POI) recommendation for users based on their history.
method SANST incorporates spatio-temporal patterns into self-attentive networks.
result SANST outperforms state-of-the-art models by up to 13.65% in nDCG@10.
Lipschitz normalization boosts deep attention models, especially for graph neural networks.
problem Gradient explosion in deep graph attention networks leads to poor performance.
method Enforcing Lipschitz continuity by normalizing attention scores.
result Deep GAT models with LipschitzNorm achieve state-of-the-art results for tasks with long-range dependencies.
Improved non-intrusive load monitoring with a novel neural network.
problem Accurately disaggregating household electricity consumption without dedicated meters.
method Developed a scale- and context-aware network with multi-scale features and contextual information.
result Significantly improved accuracy compared to state-of-the-art methods.
SANs estimate feature importance in neural models, identifying interactions and high-ranked features.
problem Understanding and interpreting black-box neural network models.
method Self-Attention Network (SAN) architecture for feature importance estimation.
result SANs identify similar high-ranked features and feature interactions as other methods, improving predictive performance.
Speech enhancement improved by adapting to unknown speakers without auxiliary signals.
problem Improving speech enhancement accuracy for unknown speakers.
method Adopting multi-task learning for speech enhancement and speaker identification, using multi-head self-attention.
result Achieved state-of-the-art performance and improved subjective quality.
MetNet forecasts precipitation up to 8 hours with high spatial and temporal resolution.
problem Precise weather forecasting for long lead times.
method Neural network architecture using axial self-attention for global context aggregation.
result MetNet outperforms Numerical Weather Prediction at forecasts of up to 8 hours.
Improved speech enhancement with MNTFA using time-frequency attention.
problem Speech enhancement with limited model size and memory.
method Designing MNTFA with self-attention modules for long sequences and joint training.
result MNTFA achieves better performance with fewer parameters than DPCRN.
EFA extends self-attention to handle mixed data types and dynamic relevance.
problem Handling high-dimensional, mixed data types with dynamic relevance.
method Probabilistic generative model using self-attention and latent factor model.
result EFA consistently outperforms existing models in complex latent structure capture and reconstruction.
This paper analyzes privacy risks in neural network pruning and proposes a defense mechanism.
problem Privacy risks in neural network pruning due to membership inference attacks.
method Investigates the impact of pruning on prediction divergence and proposes a self-attention membership inference attack.
result Proposed defense mechanism mitigates privacy risks while maintaining sparsity and accuracy.
ConViT combines CNN and ViT strengths, improving image classification.
problem Combining the strengths of CNNs and ViTs while avoiding their limitations.
method Introducing GPSA, a form of positional self-attention with a soft convolutional inductive bias.
result ConViT outperforms DeiT on ImageNet while offering improved sample efficiency.
Transformers become faster by linearizing self-attention.
problem Quadratic complexity of transformers makes them slow for long sequences.
method Expressed self-attention as a linear dot-product and used matrix product associativity to reduce complexity.
result Linear transformers are up to 4000x faster on long sequences.
SAG-VAE learns data representations and feature relations end-to-end.
problem Vanilla VAEs cannot learn relations between features.
method Inspired by Graph Neural Networks, SAG-VAE jointly infers data representations and feature relations.
result SAG-VAE generates new data via graph convolution and is robust to perturbations.
Transformer learns graph structure better with subgraph info.
problem Transformer struggles with structural similarity in graph learning.
method Structure-Aware Transformer with subgraph attention.
result Improves graph prediction benchmarks significantly.
Vision Transformers show different internal representations compared to CNNs.
problem Understanding how Vision Transformers solve image classification tasks.
method Comparative analysis of ViT and CNN architectures on image classification benchmarks.
result ViT has more uniform representations across all layers, while CNNs have more varied representations.
This work analyzes self-attention matrices using random matrix theory.
problem Understanding the theoretical behavior of self-attention layers in neural networks.
method Asymptotic spectral analysis of the attention matrix, Gaussian equivalence, and linearization.
result The singular value distribution of the attention matrix is asymptotically characterized by a linear model.
A new method improves semi-supervised learning by handling tasks with different attribute spaces.
problem Existing methods assume tasks share the same attribute space, limiting their applicability.
method Meta-learning approach that embeds labeled and unlabeled data in task-specific spaces using neural networks.
result Improves test performance on tasks with small labeled data using unlabeled and various task data.
Advanced methods of applying deep learning to structured data such as graphs have been proposed in recent years. In particular, studies have focused on generalizing convolutional neural networks to graph data, which includes redefining the convolution and the downsampling (pooling) operations for graphs. The method of …
Deep neural net solves multi-agent optimal trading problem.
problem Optimal trade execution for multiple agents and assets.
method Residual U-net with self-attention for viscosity solution approximation.
result Neural network approach outperforms finite difference methods.
Many real-world applications involve multivariate, geo-tagged time series data: at each location, multiple sensors record corresponding measurements. For example, air quality monitoring system records PM2.5, CO, etc. The resulting time-series data often has missing values due to device outages or communication errors. …