ALEXANDRIA, Va., Aug. 5 -- United States Patent no. 12,699,874, issued on Aug. 4, was assigned to GOOGLE LLC (Mountain View, Calif.).
"Leveraging redundancy in attention with reuse transformers" was invented by Venkata S. Bhojanapalli (New York), Andreas Veit (New York), Ayan Chakrabarti (New York), Frederick Liu (Bellevue, Wash.), Himanshu Jain (New York), Michal Lukasik (New York), Sanjiv Kumar (Jericho, N.Y.) and Yin-Wen Chang (New York).
According to the abstract* released by the U.S. Patent & Trademark Office: "Provided are systems and methods that improve the computational efficiency of Transformers or other attention-based neural networks or machine learning models by re-using a number of attention scores between layers and/or head...