Company Filing History:
Years Active: 2024-2025
Title: Michael A. Cogswell: Innovator in Visual Language Models
Introduction
Michael A. Cogswell is an accomplished inventor based in West Windsor, NJ (US). He has made significant contributions to the field of artificial intelligence, particularly in the development of visual language models. With a total of 3 patents to his name, Cogswell continues to push the boundaries of technology.
Latest Patents
Cogswell's latest patents include groundbreaking work in instruction-guided visual embeddings and feedback-based learning in large vision-language models. This patent outlines a method for fine-tuning a Large Visual Language Model (LVLM) by providing visual queries that consist of images and related textual queries. The LVLM processes these queries to extract visual embeddings and generates responses based on them. Additionally, he has developed a patent focused on error-based explanations for artificial intelligence behavior. This system analyzes images to output results while assigning error scores to image features, ultimately providing visual indications of these scores.
Career Highlights
Cogswell is currently employed at SRI International, where he applies his expertise in artificial intelligence and machine learning. His work has been instrumental in advancing the capabilities of AI systems, particularly in understanding and interpreting visual data.
Collaborations
Throughout his career, Cogswell has collaborated with notable colleagues such as Ajay Divakaran and Karan Sikka. These partnerships have fostered innovation and contributed to the success of various projects in the field.
Conclusion
Michael A. Cogswell is a prominent figure in the realm of artificial intelligence, with a focus on visual language models. His innovative patents and collaborative efforts continue to shape the future of technology.