Weak neural architecture search predictor patent granted
Summary
USPTO granted Microsoft Technology Licensing, LLC patent US12596908B2 for a weak neural architecture search (NAS) predictor system. The invention covers methods for iteratively searching machine learning architecture spaces using performance prediction, scoring information, and thresholding to identify optimal candidate architectures. The patent includes 20 claims covering neural network search techniques and stopping criteria.
What changed
USPTO issued patent grant US12596908B2 to Microsoft Technology Licensing, LLC for weak neural architecture search predictor technology. The patent covers systems and methods for receiving network architecture scoring information, iteratively sampling a search space by generating candidate ML architectures, learning predictors, evaluating performance, refining search spaces, and thresholding performance to identify scored output candidates. The search continues until stopping criteria such as maximum iterations or performance goals are met.
This patent grant establishes exclusive IP rights for Microsoft in the field of neural architecture search prediction. Technology companies developing AI/ML optimization tools may need to consider licensing implications or design around this patent to avoid potential infringement claims.
What to do next
- Monitor for updates
Archived snapshot
Apr 7, 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.
Weak neural architecture search (NAS) predictor
Grant US12596908B2 Kind: B2 Apr 07, 2026
Assignee
Microsoft Technology Licensing, LLC.
Inventors
Xiyang Dai, Dongdong Chen, Yinpeng Chen, Mengchen Liu, Ye Yu, Zicheng Liu, Mei Chen, Lu Yuan, Junru Wu
Abstract
A neural architecture search (NAS) with a weak predictor comprises: receiving network architecture scoring information; iteratively sampling a search space, wherein the sampling comprises: generating a set of candidate architectures within the search space; learning a first predictor; evaluating performance of the candidate architectures; and based on at least the performance of the set of candidate architectures and the network architecture scoring information, refining the search space to a smaller search space; based on at least the network architecture scoring information, thresholding the performance of candidate architectures to determine scored output candidate architectures; and reporting the scored output candidate architectures. In some examples, the candidate architectures each comprise a machine learning (ML) model, for example a neural network (NN). In some examples, searching continues to iterate until stopping criteria is met, such as a specified maximum number of iterations or a set of candidate architectures achieves a performance goal.
CPC Classifications
G06N 3/045 G06N 3/08 G06F 18/217 G06F 11/3409
Filing Date
2020-12-15
Application No.
17122428
Claims
20
Related changes
Get daily alerts for USPTO Patent Grants - 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 Grants - AI & Computing (G06N) publishes new changes.
Subscribed!
Optional. Filters your digest to exactly the updates that matter to you.