This inventor holds 1 USPTO granted patent and 7 published patent applications. Top assignee: Gdm Holding LLC. Active years: 2026.
Company Filing History:
Years Active: 2026
Title: Robert David Fergus - Innovator in Dual Encoder Training and Resource Navigation
Introduction
Robert David Fergus is an inventor based in New York, NY (US). He is known for his contributions to the field of machine learning, particularly in the development of methods and systems for training dual encoder models and resource navigation using neural networks. Although he currently holds no granted patents, his innovative ideas are reflected in his latest patent applications.
Latest Patent Applications
1. **TRAINING A DUAL ENCODER WITH A CORRECTION MODEL** - This application describes methods, systems, and apparatus for training a dual encoder model using a correction model. The approach allows for the correction of target embeddings at each training iteration without the need to explicitly recalculate each target embedding. The system involves obtaining approximated target embeddings for various target data items, processing these embeddings with a correction model to generate corrected target embeddings, and training the dual encoder model on a loss function for retrieval tasks.
2. **RESOURCE NAVIGATION USING NEURAL NETWORKS** - This application outlines methods, systems, and apparatus for resource navigation utilizing neural networks. The focus is on enhancing the efficiency and accuracy of resource navigation through advanced computational techniques.
Conclusion
Robert David Fergus is an innovative thinker whose work in dual encoder training and resource navigation showcases his commitment to advancing machine learning technologies. His latest patent applications reflect his potential to contribute significantly to the field.