paper-with-me

홈 › Papers

A Tensor Decomposition Perspective on Second-order RNNs

2024-06-07 · Maude Lizaire, Michael Rizvi-Martel, Marawan Gamal Abdel Hameed, Guillaume Rabusseau

Second-order Recurrent Neural Networks (2RNNs) extend RNNs by leveraging second-order interactions for sequence modelling. These models are provably more expressive than their first-order counterparts and have connections to well-studied models from formal language theory. However, their large parameter tensor makes computations intractable. To circumvent this issue, one approach known as MIRNN consists in limiting the type of interactions used by the model. Another is to leverage tensor decomposition to diminish the parameter count. In this work, we study the model resulting from parameterizing 2RNNs using the CP decomposition, which we call CPRNN. Intuitively, the rank of the decomposition should reduce expressivity. We analyze how rank and hidden size affect model capacity and show the relationships between RNNs, 2RNNs, MIRNNs, and CPRNNs based on these parameters. We support these results empirically with experiments on the Penn Treebank dataset which demonstrate that, with a fixed parameter budget, CPRNNs outperforms RNNs, 2RNNs, and MIRNNs with the right choice of rank and hidden size.

📄 PDF Abstract BibTeX arXiv:2406.05045

Code (1)

MaudeLiz/cprnn 공식 구현 pytorch

Tasks

Tensor Decomposition

Similar Papers 제목 키워드 기반

Long-term Forecasting using Tensor-Train RNNs

2018-01-01 · ICLR 2018 1 · Rose Yu, Stephan Zheng, Anima Anandkumar, Yisong Yue

We present Tensor-Train RNN (TT-RNN), a novel family of neural sequence architectures for multivariate forecasting in environments with nonlinear dynamics. Long-term forecasting in such systems is highly challenging, sin…

Kronecker CP Decomposition with Fast Multiplication for Compressing RNNs

2020-08-21 · Dingheng Wang, Bijiao Wu, Guangshe Zhao, Man Yao 외

Recurrent neural networks (RNNs) are powerful in the tasks oriented to sequential data, such as natural language processing and video recognition. However, since the modern RNNs, including long-short term memory (LSTM) a…

Tensor DecompositionVideo Recognition

Language Modeling Using Tensor Trains

2024-05-07 · Zhan Su, Yuqin Zhou, Fengran Mo, Jakob Grue Simonsen

We propose a novel tensor network language model based on the simplest tensor network (i.e., tensor trains), called `Tensor Train Language Model' (TTLM). TTLM represents sentences in an exponential space constructed by t…

Language ModelingLanguage Modelling

Tensor Decomposition for Compressing Recurrent Neural Network

2018-02-28 · Andros Tjandra, Sakriani Sakti, Satoshi Nakamura

In the machine learning fields, Recurrent Neural Network (RNN) has become a popular architecture for sequential data modeling. However, behind the impressive performance, RNNs require a large number of parameters for bot…

Tensor Decomposition

Compressing Recurrent Neural Networks with Tensor Ring for Action Recognition

2018-11-19 · NIPS Workshop CDNNRIA 2018 · Yu Pan, Jing Xu, Maolin Wang, Jinmian Ye 외

Recurrent Neural Networks (RNNs) and their variants, such as Long-Short Term Memory (LSTM) networks, and Gated Recurrent Unit (GRU) networks, have achieved promising performance in sequential data modeling. The hidden la…

Action RecognitionTemporal Action LocalizationTensor Decomposition