公开(公告)号：US20250103882A1

公开(公告)日：2025-03-27

申请号：US18583291

申请日：2024-02-21

Applicant: QUALCOMM Technologies, Inc.

Inventor： Dawid Jan KOPICZKO ,  Yuki Markus ASANO ,  Tijmen Pieter Frederik BLANKEVOORT

IPC: G06N3/08

Abstract: Certain aspects of the present disclosure provide techniques and apparatus for efficiently adapting a machine learning model from a base task to a downstream task based on frozen matrices. An example method generally includes receiving an input for processing through a layer of a neural network. An output of the layer of the neural network is generated based on a first product, the first product being based on a first trainable scaling vector, a first frozen matrix, a second trainable scaling vector, a second frozen matrix, and the received input.