This inventor holds 1 USPTO granted patent and 11 published patent applications. Top assignee: Qualcomm Incorporated. Active years: 2026.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2026
Title: Jaeseong You: Innovator in Parameter-Efficient Fine-Tuning
Introduction
Jaeseong You is a prominent inventor based in Seoul, South Korea. He has made significant contributions to the field of artificial intelligence, particularly in the area of large-scale models. His innovative approach to fine-tuning has garnered attention in the tech community.
Latest Patents
Jaeseong You holds a patent for "Selective parameter-efficient fine-tuning for large-scale models." This processor-implemented method involves receiving a large language model (LLM) that consists of multiple layers, each with a set of parameters. The method identifies a subset of these parameters to fine-tune for a downstream task based on a score function. An adapter is then applied to the identified subset, ensuring that only this subset is fine-tuned. This approach enhances the efficiency of model training while maintaining performance.
Career Highlights
Jaeseong You is currently employed at Qualcomm Incorporated, a leading technology company known for its advancements in mobile and wireless technology. His work at Qualcomm has allowed him to explore innovative solutions in AI and machine learning.
Collaborations
Jaeseong collaborates with talented individuals such as Sungha Choi and Jungsoo Lee. Their combined expertise contributes to the development of cutting-edge technologies in their field.
Conclusion
Jaeseong You is a notable inventor whose work in parameter-efficient fine-tuning is shaping the future of large-scale models. His contributions at Qualcomm and his innovative patent reflect his commitment to advancing technology.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
