This inventor holds 1 USPTO granted patent. Top assignee: Asapp, Inc.. 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: Barrett Martin Lattimer and His Patent on Factual-Consistency Detection
Introduction
Barrett Martin Lattimer is an inventor based in New York, New York. His patent work addresses a challenge in generative artificial intelligence: detecting when AI-generated text is inconsistent with a reference document.
Latest Patents
Lattimer is associated with 1 patent, titled “Factual-Consistency Detection Using Chunking and a Natural Language Inference Model.” The disclosed method divides a reference document into chunks containing more than one sentence, then evaluates generated text against those chunks. This approach is intended to improve accuracy and efficiency, particularly when working with long inputs. A search-tree method also helps identify relevant sentences in the reference document to support a consistency decision while using fewer model runs than earlier approaches.
Career Highlights
Lattimer works at Asapp, Inc. His patent focuses on methods for assessing the factual consistency of AI-generated text against source material.
Collaborations
The provided information names Pei-Hung Chen and Xinyuan Zhang as Lattimer’s coworkers. It does not specify whether they collaborated with him on the patent.
Conclusion
Barrett Martin Lattimer’s patent describes a chunking and search-tree approach to factual-consistency detection, aimed at making the evaluation of AI-generated text more effective for long reference documents.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
