Sparse activation-aware weight loading for ML inference

← USPTO Patent Applications

SPARSE ACTIVATION-AWARE WEIGHT LOADING AND INFERENCE FOR MACHINE LEARNING MODELS

Application US20260093965A1 Kind: A1 Apr 02, 2026

Inventors

Mahdi Heydari, Sankalp Dayal, Abhishek Sahadev Sutar, Deepak Shivarudrappa, Tariq Afzal, Rahul Bakshi

Abstract

Devices and techniques are generally described for sparse activation-aware weight loading and inference for machine learning models. In some examples, a first activation tensor may be generated for first input data. A first sparsity map may be generated for the first activation tensor. The first sparsity map may indicate respective positions of zero values and non-zero values in the first activation tensor. A first set of channels of a weight tensor that correspond to respective non-zero values from the first sparsity map may be identified. The first set of channels of the weight tensor may be loaded into memory. A machine learning model may generate output data based on the first set of channels and the first activation tensor.

CPC Classifications

G06N 3/0495

Filing Date

2024-09-30

Application No.

18901836

View original document →