Los Altos, CA, United States of America

Enxu Yan

This inventor holds 23 USPTO granted patents and 2 published patent applications, primarily in Neural Network Optimization. Top assignees: Moffett International Co., Limited, Moffett Technologies Co., Limited, Snap Inc.. Active years: 2020-2026.

IDiyas Innovation Intelligence. (2026). Inventor Profile: Enxu Yan. Retrieved from https://idiyas.com/inventor/enxu-yan

Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile

USPTO Granted Patents = 23 

% Patents Active = 95.7

Average Co-Inventor Count = 2.6

ph-index = 5

Forward Citations = 90(Granted Patents)


Company Filing History:


Years Active: 2020-2026

Loading Chart...
Areas of Expertise:
Tensor Transpose
Sparse Convolution
Neural Network Pruning
Activation Sparsity
3D Object Recognition
Few-Shot Learning
Hierarchical Weight-Sparsity
Balanced Activation
Sample Generator
Feature Maps
Domain Specific Pruning
Optimization Techniques
23 patents (USPTO):Explore Patents

Title: Enxu Yan: Innovator in Neural Network Technologies

Introduction

Enxu Yan, a prolific inventor based in Los Altos, California, holds an impressive portfolio of 19 patents to his name. His work primarily focuses on advanced methodologies in neural network pruning and optimization, contributing significantly to the field of artificial intelligence.

Latest Patents

Among his latest patents, one notable invention is the "Optimizer Based Prunner for Neural Networks." This patent describes a neural network pruning system capable of sparsely pruning neural network models using an optimizer-based approach. This innovative system operates agnostically with respect to the model architecture being pruned, by working directly on the parameter vector of the full model and the gradient vector of the loss function. The iterative updating of parameters is based on gradients while effectively zeroing out parameters according to a preconfigured penalty.

Another significant patent is the "Bank-Balanced-Sparse Activation Feature Maps for Neural Network Models." This invention details methods for generating bank-balanced-sparse activation neural network models. It involves a neural network sparsification engine that segments activation feature maps into various banks based on size and layout, ultimately increasing sparsity within each bank of the activation feature map to enhance inferencing capabilities.

Career Highlights

Throughout his career, Enxu Yan has worked with notable companies, including Moffett Technologies and Snap Inc. His innovative approaches and deep understanding of machine learning have greatly influenced the efficiency and effectiveness of neural networks in various applications.

Collaborations

Enxu has collaborated with distinguished peers such as Wei Wang and Zhibin Xiao. These partnerships underline the importance of collaborative innovation in advancing technology, especially in complex fields like artificial intelligence.

Conclusion

Enxu Yan's contributions to neural network technology exemplify the spirit of innovation. With his extensive experience and numerous patents, he continues to pave the way for future advancements in the realm of artificial intelligence, making a lasting impact on the industry.

Profile summary based on public USPTO records.
Data Sources: USPTO Patent Grant XML, Patent Center, EPO & CIPO • Normalized by IDiyas Innovation Graph. Methodology & provenance architecturePlease report any incorrect information to [email protected]
Loading…