Machine Learning Model Compression Patent Application
Summary
USPTO published patent application US20260099712A1 for Salesforce, Inc. on April 9, 2026. The application covers techniques for compressing machine learning models by removing and replacing blocks while preserving outputs, enabling execution on low-resource devices. The inventors are Romain Cosentino, Sarath Shekkizhar, Damjan Kalajdzievski, and Adam Earle.
What changed
USPTO published a new patent application from Salesforce covering machine learning model compression. The method involves receiving an ML model with multiple blocks, removing certain blocks to form an intermediate model, adding replacement blocks that generate equivalent outputs, and executing the compressed model on low-resource devices. CPC classifications include G06N 3/082 and G06N 3/0495.
For technology companies and AI developers, this patent application signals Salesforce's intellectual property position in model compression techniques for edge and low-resource deployment. Competitors developing similar compression methods should review the claims upon grant to assess potential freedom-to-operate implications.
What to do next
- Monitor for patent grant or office action responses
Archived snapshot
Apr 11, 2026GovPing captured this document from the original source. If the source has since changed or been removed, this is the text as it existed at that time.
MACHINE LEARNING MODEL COMPRESSION
Application US20260099712A1 Kind: A1 Apr 09, 2026
Assignee
Salesforce, Inc.
Inventors
Romain Cosentino, Sarath Shekkizhar, Damjan Kalajdzievski, Adam Earle
Abstract
Techniques are described herein for a method of machine learning model compression. The method includes receiving a machine learning model comprising a plurality of blocks. The method further includes removing one or more blocks of the plurality of blocks to obtain an intermediate machine learning model comprising a subset of the plurality of blocks. The method further includes adding a block to the intermediate machine learning model to obtain a compressed machine learning model. The block generates an output corresponding to an output of the removed one or more blocks of the plurality of blocks. The method further includes executing the compressed machine learning model on a low resource device.
CPC Classifications
G06N 3/082 G06N 3/0495
Filing Date
2024-10-03
Application No.
18905761
Related changes
Get daily alerts for USPTO Patent Applications - AI & Computing (G06N)
Daily digest delivered to your inbox.
Free. Unsubscribe anytime.
Source
About this page
Every important government, regulator, and court update from around the world. One place. Real-time. Free. Our mission
Source document text, dates, docket IDs, and authority are extracted directly from USPTO.
The summary, classification, recommended actions, deadlines, and penalty information are AI-generated from the original text and may contain errors. Always verify against the source document.
Classification
Who this affects
Taxonomy
Browse Categories
Get alerts for this source
We'll email you when USPTO Patent Applications - AI & Computing (G06N) publishes new changes.
Subscribed!
Optional. Filters your digest to exactly the updates that matter to you.