The workflow of the automatically differentiable tensor network method for compressing neural networks. (IMAGE)

Intelligent Computing

The workflow of the automatically differentiable tensor network method for compressing neural networks.

Caption

(A) Illustration of a convolutional neural network (NN) whose variational parameters (T) are encoded in the automatically differentiable tensor network (ADTN) shown in (B). The ADTN contains many fewer parameters than T.

Credit

Yong Qing et al.

Usage Restrictions

Credit must be given to the creator.

License

CC BY

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.