Questa è la nostra ultima selezione di pubblicazioni e brevetti in inglese su Large Language Models (LLM), tra numerose riviste scientifiche online, classificate e incentrate su large language model, LLM, generative pre-trained transformer, pre-training, transformer architecture, gradient descent, GPT, tokenization, generative model, self-attention mechanism, masked language model e MLM.
Method, apparatus, device, medium and product for simulating production facility
Patent published on the 2026-05-28 in WO under Ref WO2026107811 by QIAN YINGJIE [CN] (Lv Qizhen [cn], Zhu Tongshuai [cn], Qian Yingjie [cn])
Abstract: A method, an apparatus, an electronic device, a computer-readable storage device, and a computer program product for simulating a production facility are provided. The method comprises receiving a user input to simulate the production facility from a user of an industrial simulation software, wherein the user input comprises one or more simulation requirements and a type of the production facility. The method further comprises determining, via a generative model associated with the industrial si[...]
Our summary: The content describes a method and apparatus for simulating a production facility. It involves receiving user input to define simulation requirements and facility type. A generative model is used to determine the facility layout and simulate its performance parameters.
simulation, production facility, generative model, performance parameters
Patent
Communication method and apparatus
Patent published on the 2026-05-28 in WO under Ref WO2026108188 by HUAWEI TECH CO LTD [CN] (Zhou Zhi [cn], He Jia [cn], Qiu Jing [cn])
Abstract: Provided in the present application are a communication method and apparatus, which are capable of acquiring radio frequency channel map data. The method comprises: a first communication apparatus sending first information to a second communication apparatus, wherein the first information is used for requesting the acquisition of second information, the second information comprises information of a generative model and/or radio frequency channel map data, and the generative model is used for gen[...]
Our summary: The communication method involves a first apparatus requesting channel map data from a second apparatus. The second apparatus provides information related to a generative model and radio frequency channel map data. The first apparatus then receives this second information for further processing.
communication method, apparatus, radio frequency, generative model
Patent
Systems and methods for optimized fingerprinting and tokenization in server drift analysis
Patent published on the 2026-05-28 in US under Ref US20260149639 by BANK OF AMERICA CORP [US] (Alizzi Jason A [us], Lozes John Andres [us], Yu Chia-ho [us], Maniotis Daniel M [us], Bosi Robert [us], Jain Amit [us], Misra Pravesh K [us], Carlson Kevin Roy [us], Simon Matthew E [us], Badawy Tamer [us], Nair Sreeni R [us], Gupta Dharmendra Kumar [in],)
Abstract: Systems, computer program products, and methods are described herein for optimized fingerprinting and tokenization in server drift analysis. The present disclosure is configured to detect configuration drift across entire computer servers by utilizing distributed agents that collect and analyze metadata such as file modification dates, sizes, and cryptographic hashes. These agents sanitize sensitive information and conditionally tokenize environment-specific values, such as server names, IP addr[...]
Our summary: The system optimizes fingerprinting and tokenization for server drift analysis. It utilizes distributed agents to collect and sanitize metadata for accurate configuration comparisons. This automation enhances efficiency, accuracy, and security in large, dynamic environments.
Fingerprinting, Tokenization, Server Drift, Metadata
Patent
Software bug analysis and resolution using large language models
Patent published on the 2026-05-28 in US under Ref US20260147689 by SAP SE [DE] (Li Hui [cn])
Abstract: [0000] Methods, systems, and computer-readable storage media for debugging errors occurring in software and can include actions of receiving a current bug ticket (CBT) including bug data descriptive of a bug occurring within a software system, prompting a LLM using a first prompt to return a first embedding representative of at least a portion of the CBT, determining a set of candidate historical bug tickets (HBTs) using the first embedding and a set of HBT embeddings, prompting the LLM using a [...]
Our summary: This method involves receiving a current bug ticket and utilizing a large language model to generate embeddings. It identifies candidate historical bug tickets based on these embeddings. The process updates the current bug ticket with similar historical bug tickets for resolution.
bug analysis, large language models, software debugging, historical bug tickets
Patent
Performance–based content generation and exploration using multimodal generative model
Patent published on the 2026-05-28 in US under Ref US20260148044 by ADOBE INC [US] (Yan Zhenyu [us], Mitra Saayan [us], Sinha Ritwik [us], Koh Eunyee [us], Faieta Baldo [us], Swaminathan Viswanathan [us])
Abstract: [0000] Some aspects relate to technologies for performance-guided content generation and exploration using a multimodal generative model with a joint latent space learned from digital content items and their corresponding performance metrics. In accordance with some aspects, input is received for content generation. The input includes a digital content item and is encoded by one or more encoders of the multimodal generative model into a latent representation in the joint latent space. A latent s[...]
Our summary: This content discusses a multimodal generative model that uses a joint latent space to guide content generation based on performance metrics. It describes the process of encoding digital content into a latent representation and transforming it for output generation. Additionally, the model predicts performance metrics for the generated content.
multimodal generative model, performance metrics, latent space, content generation
Patent
Automated data extraction using large language model
Patent published on the 2026-05-28 in US under Ref US20260147995 by PAYPAL INC [US] (Arulmozhi Suraj [in], Oppenheim Hagar [us], Soni Trishul [us], Modaresi Sina [us], Verma Kunal [in], Palanisamy Senthil Sugesh [in], Shukla Aditya [in], Hochma Gal [il], Rozenker Mor [il], Marad Zohar Li [il], Wang Ruobai [cn], Ge Yirou [cn], Subash Ashok)
Abstract: [0000] Techniques are disclosed relating to extracting data from a document, using a large language model (LLM), to populate fields in a data structure. A computer system may receive a request to populate multiple fields of a data structure with data extracted from text of a document. The computer system parses the text using an LLM (as well as regular expressions or other parsing techniques in some embodiments). The parsing includes issuing, to the LLM, a sequence of queries targeting individua[...]
Our summary: The system automates data extraction from documents using a large language model. It parses text to populate fields in a data structure through targeted queries. A validation algorithm ensures the accuracy of the extracted data, and OCR may be employed to convert document images to text.
data extraction, large language model, validation algorithm, optical character recognition
Patent
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
capability-based monitoring
Published on 2026-05-15 by Katherine C. Kellogg, Bingyang Ye, Yifan Hu, Guergana K. Savova, Byron Wallace, Danielle S. Bitterman @NATURE npj
Abstract: npj Digital Medicine, Published online: 15 May 2026; doi:10.1038/s41746-026-02740-0Large language models (LLMs) have been rapidly adopted in healthcare, but oversight strategies are lacking. We propose capability-based monitoring, motivated by the fact that LLMs are generalist systems whose overlapping internal capabilities are reused across numerous downstream tasks. This approach organizes monitoring around shared capabilities to enable cross-task detection of systemic weaknesses, long-tail er[...]
Our summary: Capability-based monitoring addresses oversight gaps in large language models used in healthcare. It organizes monitoring around shared capabilities to detect systemic weaknesses and errors. The approach provides guidance for developers, organizational leaders, and policymakers.
capability-based monitoring, large language models, healthcare oversight, systemic weaknesses
Publication











