This inventor holds 1 USPTO granted patent. Top assignee: Ford Global Technolgoies, LLC. Active years: 2026.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2026
Title: Diana Mann and Length-Based Large Language Models.
Introduction
Diana Mann is an inventor based in Herndon, Virginia. Her patent work addresses a practical challenge in large language models: generating text that meets a requested length.
Latest Patents
Mann is associated with one patent, titled “Length-Based Large Language Models.” The invention describes a computer with a processor and memory containing instructions for receiving a prompt that includes input text and a target length. A length-guidance embedding vector encodes the target size, enabling the model to generate output with a word count within a user-determined tolerance of that target.
Career Highlights
Mann is associated with Ford Global Technologies, LLC. Her patent reflects work on controlling the length of text generated by large language models.
Collaborations
Lucas Ross is identified as Mann’s coworker.
Conclusion
Diana Mann’s patent focuses on making large language model outputs more predictable in length, connecting text-generation technology with a clearly specified user requirement.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
