paper-with-me

홈 › Papers

One-Shot Pruning of Recurrent Neural Networks by Jacobian Spectrum Evaluation

2019-11-30 · ICLR 2020 1 · Matthew Shunshi Zhang, Bradly Stadie

Recent advances in the sparse neural network literature have made it possible to prune many large feed forward and convolutional networks with only a small quantity of data. Yet, these same techniques often falter when applied to the problem of recovering sparse recurrent networks. These failures are quantitative: when pruned with recent techniques, RNNs typically obtain worse performance than they do under a simple random pruning scheme. The failures are also qualitative: the distribution of active weights in a pruned LSTM or GRU network tend to be concentrated in specific neurons and gates, and not well dispersed across the entire architecture. We seek to rectify both the quantitative and qualitative issues with recurrent network pruning by introducing a new recurrent pruning objective derived from the spectrum of the recurrent Jacobian. Our objective is data efficient (requiring only 64 data points to prune the network), easy to implement, and produces 95% sparse GRUs that significantly improve on existing baselines. We evaluate on sequential MNIST, Billion Words, and Wikitext.

📄 PDF Abstract BibTeX arXiv:1912.00120

Code (0)

등록된 구현이 없습니다.

Tasks

Network Pruning

Methods 이 논문이 사용한 방법론

Pruning 설명 없음
Sigmoid Activation 설명 없음
Tanh Activation 설명 없음
GRU A Gated Recurrent Unit, or GRU, is a type of recurrent neural network. It is similar to an LSTM, but only has two gates - a reset…
LSTM An LSTM is a type of recurrent neural network that addresses the vanishing gradient problem in vanilla…

Similar Papers 제목 키워드 기반

Almost Sure Asymptotic Freeness of Neural Network Jacobian with Orthogonal Weights

2019-08-11 · Tomohiro Hayase

A well-conditioned Jacobian spectrum has a vital role in preventing exploding or vanishing gradients and speeding up learning of deep neural networks. Free probability theory helps us to understand and handle the Jacobia…

Hyperpruning: Efficient Search through Pruned Variants of Recurrent Neural Networks Leveraging Lyapunov Spectrum

2025-06-09 · Caleb Zheng, Eli Shlizerman

A variety of pruning methods have been introduced for over-parameterized Recurrent Neural Networks to improve efficiency in terms of power consumption and storage utilization. These advances motivate a new paradigm, term…

Hyperparameter OptimizationNetwork Pruning

Multi-Agent Actor-Critic with Harmonic Annealing Pruning for Dynamic Spectrum Access Systems

2025-03-19 · George Stamatelis, Angelos-Nikolaos Kanatas, George C. Alexandropoulos

Multi-Agent Deep Reinforcement Learning (MADRL) has emerged as a powerful tool for optimizing decentralized decision-making systems in complex settings, such as Dynamic Spectrum Access (DSA). However, deploying deep lear…

Decision MakingDeep Reinforcement LearningNetwork Pruning

Gradients are Not All You Need

2021-11-10 · Luke Metz, C. Daniel Freeman, Samuel S. Schoenholz, Tal Kachman

Differentiable programming techniques are widely used in the community and are responsible for the machine learning renaissance of the past several decades. While these methods are powerful, they have limits. In this sho…

All

The Emergence of Spectral Universality in Deep Networks

2018-02-27 · Jeffrey Pennington, Samuel S. Schoenholz, Surya Ganguli

Recent work has shown that tight concentration of the entire spectrum of singular values of a deep network's input-output Jacobian around one at initialization can speed up learning by orders of magnitude. Therefore, to …