This inventor holds 2 USPTO granted patents and 2 published patent applications and 1 EPO patent. Top assignee: Google Inc.. Active years: 2023-2024.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2023-2024
Title: Sameer Kumar - Innovator in Neural Network Training
Introduction
Sameer Kumar is a prominent inventor based in Mountain View, CA (US). He has made significant contributions to the field of machine learning, particularly in the training of neural networks. With a total of 2 patents, his work focuses on enhancing the efficiency and effectiveness of distributed training methods.
Latest Patents
Sameer Kumar's latest patents include innovative methods and systems for training neural networks. One of his patents, titled "Training neural networks using distributed batch normalization," outlines techniques for performing reduction of gradient vectors during distributed training. This method involves receiving batches at multiple devices, performing forward passes, and applying batch normalization using distributed means and variances. Another patent, "Cross replica reduction on networks having degraded nodes," describes a method for training machine learning models across nodes in an n-dimensional network topology, ensuring effective gradient vector reduction even in the presence of degraded nodes.
Career Highlights
Sameer Kumar is currently employed at Google Inc., where he continues to push the boundaries of machine learning technology. His work has been instrumental in developing advanced algorithms that improve the performance of neural networks in various applications.
Collaborations
Throughout his career, Sameer has collaborated with notable colleagues, including Bjarke Hammersholt Roune and Norman Paul Jouppi. These collaborations have further enriched his research and contributed to the advancement of machine learning methodologies.
Conclusion
Sameer Kumar's innovative work in neural network training exemplifies the impact of dedicated inventors in the field of technology. His patents reflect a commitment to improving machine learning processes, making significant strides in the industry.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
