This is our latest selection of worldwide publications and patents in english on Large Language Models (LLM), between many scientific online journals, classified and focused on large language model, LLM, generative pre-trained transformer, pre-training, transformer architecture, gradient descent, GPT, tokenization, generative model, self-attention mechanism, masked language model and MLM.
Systems and methods for using client system intelligence for distributed service system configuration
Patent published on the 2026-06-11 in US under Ref US20260163813 by STRIPE INC [US] (Gehman Samuel Ishmael [us], Jermsurawong Jermsak [us], Vafeias Efstathios [us], Brown Jonathon Daniel [us], Kedia Gautam [us])
Abstract: A method and apparatus for executing a service of a server computer system for a client system are described. The method may include defining a client system descriptor file that includes a set of signals suitable for input into a large language machine learning model (LLM), and collecting data associated with a first client system that is representative of the set of signals. The data can then be compressed into the set of signals for a first client system descriptor file allocated for the clie[...]
Our summary: The method involves defining a client system descriptor file with signals for a large language model. Data from a client system is collected and compressed into this descriptor file. A query is executed using the descriptor file, leading to actions by the server s service based on the model s output.
client system intelligence, distributed service configuration, machine learning model, client system descriptor file
Patent
Systems and methods for increasing situational awareness of alerts generated by a video monitoring system
Patent published on the 2026-06-11 in US under Ref US20260162431 by HONEYWELL INT INC [US] (Gupta Anchal [in], Eswara Lalitha M [in])
Abstract: Methods and systems are provided for increasing situational awareness of alerts from a video monitoring system. Video analytics algorithms detect conditions in a video stream and generate alerts. For each alert, a video clip is extracted containing frames before and/or after the alert. A Generative AI video-to-text summarization model generates text summaries of the video frames, which are processed by a Large Language Model to generate context for each alert. Enhanced alerts are output containi[...]
Our summary: The system enhances situational awareness by generating context for alerts from video monitoring. Video analytics detect conditions and produce alerts, with video clips extracted for analysis. A generative AI model creates text summaries, which are processed to improve alert information and support future predictions.
video analytics, situational awareness, generative AI, alert processing
Patent
Code infraction quality validator with llm proposed solutions
Patent published on the 2026-06-11 in US under Ref US20260161530 by SERVICENOW INC [US] (Dyson-coope Alexander Pierre [gb])
Abstract: [0000] A method includes obtaining a portion of a code file that is associated with metadata and selecting a prompt template from a set of prompt templates. Each of the prompt templates includes a respective request and a respective metadata placeholder. The method includes populating the respective metadata placeholder of the prompt template with the metadata associated with the portion of the code file. The method includes generating a code modification based on the portion of the code file, t[...]
Our summary: This method involves obtaining a code file portion linked to metadata. It selects a prompt template and populates it with the relevant metadata. Finally, it generates and applies a code modification based on the populated template.
code validation, metadata processing, prompt templates, code modification
Patent
Method and apparatus for modifying architecture of large language model
Patent published on the 2026-06-11 in US under Ref US20260161945 by LG ELECTRONICS INC [KR] (Byun Sanghyun [us], Gou Jimmy [us], Yildiz Bahattin [us], Malpeddi Nilesh [us], De Micheli Gabrielle [us], Chung Wooseong [us], Song Jacob [us])
Abstract: [0000] According to at least one embodiment, a computer-implemented method of modifying an architecture of a large language model (LLM) includes compressing an embedding layer of the LLM to reduce a size of a parameter space of the LLM, wherein the embedding layer has an embedding dimension of n, wherein compressing the embedding layer includes utilizing a first intermediate mapping configured to map a token to an m-dimensional vector, and wherein m is less than n. The method further includes co[...]
Our summary: The method modifies the architecture of a large language model by compressing its embedding layer and transformer layers. This compression reduces the size of the model s parameter space. It utilizes a mapping to convert tokens into lower-dimensional vectors.
large language model, embedding layer, parameter space, transformer layers
Patent
Enhancing a code base using dynamic retrieval-augmented generation (rag) with run-time prompt enrichment
Patent published on the 2026-06-11 in US under Ref US20260161371 by ZSCALER INC [US] (Venkatesh Golla Sai [in], Shyju Saurav [in])
Abstract: [0000] Systems and methods for enhancing software code are provided. A method, according to one implementation, includes receiving a code base developed by one or more software developers and receiving a prompt for requesting enhancement to the code base. Also, the method includes a step of using a dynamic Retrieval-Augmented Generation (RAG) component and a Knowledge Base (KB) repository to enrich the prompt. Based on the enriched prompt, the method further includes a step of using a Large Lang[...]
Our summary: The method enhances software code by using a dynamic RAG component to enrich prompts. It integrates a Knowledge Base repository for improved context. A Large Language Model tool is then employed to enhance the code base effectively.
dynamic retrieval-augmented generation, run-time prompt enrichment, Knowledge Base, Large Language Model
Patent
Systems and methods for narrative generation from embeddings
Patent published on the 2026-06-11 in US under Ref US20260162169 by ROYAL BANK OF CANADA [CA] (Tung Frederick [ca], Deng Ruizhi [ca], Zhao He [ca])
Abstract: [0000] Systems and methods for narrative generation. Data representations in a first data domain are received. The data representations of the first data domain are transformed into data representations of a second data domain by an adapter. Narratives corresponding to the data representations can be generated using a large language model by interpreting the data representations of the second data domain.[...]
Our summary: This document describes systems and methods for generating narratives from data embeddings. It involves transforming data representations from one domain to another using an adapter. A large language model is then utilized to create narratives based on the transformed data representations.
narrative generation, embeddings, data transformation, large language model
Patent
Enhancing precision in question-answering for FDM troubleshooting from technical documents
Published on 2026-05-27 by @OXFORD
Abstract: AbstractWith accelerating digital transformation in manufacturing, large language models (LLMs) are increasingly being considered for leveraging domain-specific knowledge for decision-making and problem-solving. However, LLMs are prone to exhibiting hallucination when dealing with external information or specialized domain knowledge that falls beyond their training scope. To mitigate this issue, a retrieval-augmented generation (RAG) approach that incorporates external documents is used. However[...]
Our summary: This study introduces Keyword-RAG, a novel method that enhances precision in question-answering for FDM troubleshooting. It integrates domain-specific keyword extraction to maintain semantic coherence during retrieval. Experimental results show significant improvements in context recall and precision, benefiting decision-making in manufacturing environments.
Keyword-RAG, retrieval-augmented generation, domain-specific knowledge, precision enhancement
Publication
Expert perspectives on the ecosystem of medical AI oversight in the GenAI era
Published on 2026-05-22 by Stephen Gilbert, Tinglong Dai @NATURE npj
Abstract: npj Digital Medicine, Published online: 22 May 2026; doi:10.1038/s41746-026-02785-1In this third article of a three-article series on interviews with Dr Shantanu Nundy, an advisor in the office of the Commissioner of the US FDA on AI governance and digital health policy, we explore how the oversight of generative AI (GenAI) and large language model (LLM)-enabled medical devices extends beyond the remit of any single regulator, requiring coordination among government agencies, health systems, and[...]
Our summary: The article discusses the complexities of overseeing medical AI in the GenAI era. It emphasizes the need for collaboration among various stakeholders, including regulators and health systems. Additionally, it provides practical guidance for innovators on engaging with the FDA.
medical AI, oversight, generative AI, FDA
Publication











