This inventor holds 1 USPTO granted patent. Top assignee: Intuit, Inc.. Active years: 2026.
Company Filing History:
Years Active: 2026
Title: Yiheng Shu: Innovator in Large Language Model Memory
Introduction
Yiheng Shu is a notable inventor based in Mountain View, CA (US). He has made significant contributions to the field of artificial intelligence, particularly in the development of large language models. His innovative work focuses on enhancing memory capabilities for episodic and temporal reasoning.
Latest Patents
Yiheng Shu holds a patent for a groundbreaking invention titled "Large language model memory for episodic and temporal reasoning." This patent involves a method that includes obtaining episodic text describing a set of atomic events. It generates a gist statement from the episodic text in natural language for each atomic event. The process also involves extracting a fact triple from the episode text and gist list, which comprises a subject, predicate, and object. Furthermore, it updates a graph in storage with a set of fact nodes and establishes connections between these nodes based on the predicate. The method links the gist node to the set of fact nodes and utilizes the graph to respond to queries. Yiheng Shu's patent is a testament to his innovative approach in the realm of artificial intelligence.
Career Highlights
Yiheng Shu is currently employed at Intuit, Inc., where he continues to push the boundaries of technology and innovation. His work at Intuit has allowed him to collaborate with other talented professionals in the field.
Collaborations
Some of Yiheng Shu's coworkers include Xiang Gao and Padmaja Jonnalagedda. Their collaborative efforts contribute to the advancement of technology and innovation within their organization.
Conclusion
Yiheng Shu is a prominent inventor whose work in large language model memory has the potential to revolutionize how machines understand and process information. His contributions to the field of artificial intelligence are noteworthy and continue to inspire future innovations.