US Patent Issued to Salesforce on Oct. 21 for "Extractive-abstractive large language model summarization with farthest point sampling" (Washington, California Inventors)

ALEXANDRIA, Va., Oct. 28 -- United States Patent no. 12,450,273, issued on Oct. 21, was assigned to Salesforce Inc. (San Francisco).

"Extractive-abstractive large language model summarization with farthest point sampling" was invented by Bin Bi (Redmond, Wash.), Shiva Kumar Pentyala (Irvine, Calif.), Sitaram Asur (Palo Alto, Calif.), Na Cheng (Yarrow Point, Wash.) and Zhichao Wang (San Jose, Calif.).

According to the abstract* released by the U.S. Patent & Trademark Office: "In some systems, a set of sentences of a relatively large document may be vectorized into a set of vectors via an embedding model for summarization. Further, a subset of vectors of the set of vectors may be selected via a farthest point sampling (FPS) procedure based ...