Online speech recognition, where the model only accesses context to the ...
Recurrent Neural Networks (RNNs) offer fast inference on long sequences ...
Structured state space sequence (S4) models have recently achieved
state...
Performant Convolutional Neural Network (CNN) architectures must be tail...
Transformers have been essential to pretraining success in NLP. Other
ar...
Visual data such as images and videos are typically modeled as
discretiz...
Linear time-invariant state space models (SSM) are a classical model fro...
State space models (SSM) have recently been shown to be very effective a...
The use of Convolutional Neural Networks (CNNs) is widespread in Deep
Le...
Developing architectures suitable for modeling raw audio is a challengin...
A central goal of sequence modeling is designing a single principled mod...
Recurrent neural networks (RNNs), temporal convolutions, and neural
diff...
This paper studies Principal Component Analysis (PCA) for data lying in
...
Modern neural network architectures use structured linear transformation...
In real-world classification tasks, each class often comprises multiple
...
Similarity-based Hierarchical Clustering (HC) is a classical unsupervise...
A central problem in learning from sequential data is representing cumul...
Classifiers in machine learning are often brittle when deployed. Particu...
Gating mechanisms are widely used in neural network models, where they a...
In this paper we consider the following sparse recovery problem. We have...
Fast linear transforms are ubiquitous in machine learning, including the...
The low displacement rank (LDR) framework for structured matrices repres...
Hyperbolic embeddings offer excellent quality with few dimensions when
e...
Data augmentation, a technique in which a training set is expanded with
...