Reut Moshe

Title: Reut Moshe: Innovator in Homomorphic Encryption Deep Learning Architectures

Introduction

Reut Moshe is a prominent inventor based in Tel Aviv, Israel. She has made significant contributions to the field of deep learning and homomorphic encryption. Her innovative work focuses on optimizing deep learning models for secure data processing.

Latest Patents

Reut Moshe holds a patent titled "Self-attention in homomorphic encryption deep learning architectures." This patent presents mechanisms for optimizing a deep learning computer model for homomorphic encryption workload processing. The mechanisms involve modifying an original deep learning model by replacing a self-attention layer with an HE-friendly self-attention layer that utilizes a Power SoftMax function without exponent terms. This results in a modified deep learning model architecture that is trained to execute efficiently on HE workloads.

Career Highlights

Reut Moshe is currently employed at International Business Machines Corporation (IBM). Her work at IBM involves cutting-edge research and development in the area of machine learning and encryption technologies. She has demonstrated a strong commitment to advancing the capabilities of deep learning systems.

Collaborations

Reut collaborates with notable colleagues, including Allon Adir and Ramy Masalha. These partnerships enhance her research and contribute to the innovative projects at IBM.

Conclusion

Reut Moshe is a trailblazer in the field of homomorphic encryption and deep learning. Her patent and work at IBM reflect her dedication to innovation and excellence in technology.