Alerta

OPTIMIZING ARTIFICIAL INTELLIGENCE GENERATED CODE ON THE COMPUTING CONTINUUM

Absstract of: WO2026173886A1

Systems and methods for optimizing artificial intelligence generated code on the computing continuum. An instruction code that instructs a machine learning model (MLM) can be modified (510) to obtain a modified instruction code that incorporates a dynamic control flow that limits processing of software code based on a target time period. One or more candidate codes based on the modified instruction code can be generated (520). One or more service paths within the dynamic control flow of the one or more candidate codes can be executed (530) to obtain an optimized generated code having detailed responses within a threshold for downstream tasks by asynchronously performing sub-tasks from the one or more service paths to optimal computing nodes.

MAIA Outcome Feedback Computing System

Absstract of: US20260245686A1

0000 MAIA Outcome Feedback Computing System provides users with a real-time, document guidance, interface to upload medical claims, in multiple different formats. Document classification is performed on the uploaded data to determine a document type. Claim features are identified and extracted specific to each contextual document type, forwarded to a pre-approval and feedback manager and used to perform semantic and keyword searches on knowledge databases specific to each document type. The claims, and relevant data are input into a machine learning model, trained to predict medical billing codes using medical claims and generate: a confidence score and results summary for each billing code. Based on a comparison of the confidence score to a threshold, one of a plurality of validation processes is performed on each billing code, results are output to a user. Validated codes may be automatically submitted to third-party insurance providers, monitored for denials, and automatically appealed.

AI INFERENCING TO DETECT ANOMALIES IN IoT SYSTEMS

Absstract of: US20260244953A1

0000 Systems and methods are disclosed herein for monitoring an Internet of Things (IoT) platform. The systems and methods can obtain stack data, profile data, logs, or metrics that are associated with the IoT platform and extract features from each of the stack data, profile data, logs, or metrics. The extracted features from each of the stack data, profile data, logs, or metrics can be input into a separate machine learning model that produces outputs. The outputs can be input into an aggregator to determine whether the IoT platform is anomalous based on the outputs. When at least one output has an anomaly, the systems and methods can determine that the IoT platform is anomalous. The systems and methods can include false positive checks can be performed to ensure accuracy of the determination.

TIME-CONTINUOUS ANNOTATION PROCESSING SYSTEM AND METHODS FOR USE THEREWITH

Absstract of: US20260244991A1

A time-continuous annotation processing system operates by: sending a segment of time-continuous audiovisual (A/V) data to a first plurality of client devices; receiving subjective time-continuous annotation data corresponding to the segment of time-continuous A/V data, the subjective time-continuous annotation data indicating time-continuous annotated values of a subjective parameter varying over a time period of the segment of time-continuous A/V data; determining, via a subjective gold standard analysis tool, when there is a single ordinal agreement between the subjective time-continuous annotation data received from the first plurality of client devices, and in response: generating, via the subjective gold standard analysis tool, a first ordinal time-continuous gold standard annotation based on the ordinal agreement between the subjective time-continuous annotation data received from the first plurality of client devices; and constructing, via an A/V training data set post processing tool, an annotated time-continuous A/V training dataset to include the first ordinal time-continuous gold standard annotation and the segment of time-continuous A/V data; determining, via the subjective gold standard analysis tool, when there are a plurality of differing ordinal agreements of the subjective time-continuous annotation data received from the first plurality of client devices, and in response: generating, via the subjective gold standard analysis tool, a first plurality of diffe

METHODS AND MECHANISMS FOR USING MACHINE LEARNING TO GENERATE DIAGNOSTIC REPORTS BASED ON ANOMALOUS DATA

Absstract of: WO2026173646A1

A system configured to detect, by a processing device, an anomaly that occurred during a manufacturing process performed by a substrate processing system. A set of prompts is generated based on the anomaly. Each prompt is correlated to specific data obtained from one or more datastores. For each prompt, a respective output of the first trained machine learning model is obtained and a structured prompt is generated based on respective outputs. The structured prompt is provided as input to a second trained machine learning model an output of the second trained machine learning model is obtained. The output of the second trained machine learning model comprising a diagnostic report associated with the anomaly.

METHODS AND SYSTEMS FOR PREDICTING SEQUENTIAL ACTIONS USING DEEP GENERATIVE MODELS

Absstract of: US20260244936A1

Methods and systems for implementing an action prediction framework associated with a user are described. An action prediction model generates a plurality of synthetic action sequences and corresponding synthetic end states for the user, based on a sequence of historical actions and corresponding historical end states. A plurality of pathways are projected through the plurality of synthetic action sequences, for assisting the user in arriving at a desired end state. During a training phase, a machine learning model is trained to learn a plurality of implicit features related to user behavior, for generating the synthetic action sequences and pathways. During an inference phase, the action prediction framework identifies waypoints associated with recommended user or system actions for assisting the user in reaching the end state more efficiently. The disclosed methods and systems may enable robust and efficient sequential action prediction while minimizing resource consumption associated with computationally expensive foundation models.

SYSTEMS AND METHODS FOR TRAINING MACHINE LEARNING MODELS USING FORWARD VELOCITIES

Absstract of: WO2026173714A1

A computer system for labeling anomalous data for re-training a scoring machine-learning model is provided. The computer system includes a processor programmed to: receive transaction data associated with a plurality of declined transactions; apply a scoring model to the transaction data for the plurality of declined transactions; rank the plurality of declined transactions from low probability to high probability of fraud; apply a labeling model to the transaction data of a set of the plurality of declined transactions, the set including a batch of the declined transactions having higher probability scores assigned thereto; generate, using the labeling model, a precision percentage for the set of the plurality of declined transactions representing a ratio of the declined transactions labeled as fraud by the labeling model relative to the total number of declined transactions included in the set of declined transactions; and refine the precision percentage by examining subsets of the set.

AGENTIC DATA FABRIC WITH PREDICTIVE INTELLIGENCE

Absstract of: WO2026174258A1

Generating profiles that consolidate tenant data with interaction and transaction records from heterogeneous data sources in real time. Executing AI agents including an orchestration agent that orchestrates other agents. Providing a model interface layer that securely connects to an external machine learning model of a third-party agent while enforcing data governance rules, wherein the machine learning model can securely access customer data from the profiles of the multi-tenant platform under contextaware policies. Providing secure access to data from the profiles to the model through the model interface layer without exporting the data into a separate repository, such that the model processes live enterprise data in place. Receiving a predictive output derived from the exported data. Updating a profile by writing the predictive output as a new attribute of that profile, thereby enriching the profile with machine-generated insights in real time to create an enriched profile.

Automated analytical content generation using large language models

Absstract of: AU2025271014A1

Aspects of the present disclosure relate to automated analytical content generation. Embodiments include receiving data from one or more data sources. Embodiments further include extracting trends from the data using a heuristic algorithm. Embodiments further include providing an input based on the extracted trends to a generative machine learning model that has been configured to generate content based on extracted trends. Embodiments further include receiving, from the generative machine learning model based on the input, content that represents the extracted trends. Embodiments further include displaying the content via a user interface. ov o v RECEIVE DATA FROM ONE OR MORE DATA SOURCES EXTRACT TRENDS FROM THE DATA USING A HEURISTIC ALGORITHM PROVIDE AN INPUT BASED ON THE EXTRACTED TRENDS TO A GENERATIVE MACHINE LEARNING MODEL THAT HAS BEEN CONFIGURED TO GENERATE CONTENT BASED ON EXTRACTED RECEIVE, FROM THE GENERATIVE MACHINE LEARNING MODEL BASED ON THE INPUT, CONTENT THAT REPRESENTS THE EXTRACTED DISPLAY THE CONTENT VIA A USER INTERFACE RECEIVE DATA FROM ONE OR MORE DATA SOURCES ov o v

METHOD FOR OPERATING A BEAM DEVICE, COMPUTER PROGRAM AND BEAM DEVICE FOR CARRYING OUT THE METHOD AS WELL AS A METHOD FOR GENERATING A TRAINING DATA SET AND A METHOD FOR TRAINING A MACHINE LEARNING MODEL

Absstract of: US20260246526A1

The system described herein relates to operating a beam device for obtaining information about an object. Moreover, the invention relates to a computer program product having a program code, which, when executed, controls the beam device in such a way that the method for operating the beam device is carried out. Additionally, the invention relates to a method for generating a training data set for a processing unit and/or for a machine learning model. Furthermore, the invention relates to a method for training a machine learning model of a beam device. The processing unit determines which machine learning model of a plurality of machine learning models is to be used for determining control values of control parameters. The control values of the control parameters are used to operate the control unit for generating the information about the object.

APPARATUS AND METHODS FOR GENERATING STRUCTURED DATA OUTPUTS

Absstract of: US20260244949A1

0000 Apparatus for generating structured data outputs and methods used therein include a processor and a memory connected to the processor, wherein the memory contains instructions configuring the processor to receive entity data associated with an entity, the entity data including projection data and location-based data, determine at least a selection criterion as a function of the entity data, receive from a data repository a plurality of metrics as a function of the at least a selection criterion, select at least an output parameter by applying the at least a selection criterion to a plurality of output parameters, as a function of the plurality of metrics, and synthesize, using an output generation machine-learning model trained on output generation training data, a structured data output as a function of the at least an output parameter, wherein the structured data output includes a plurality of event handler graphics.

TRAINING METHOD AND ELECTRONIC DEVICE FOR QUANTUM MACHINE LEARNING

Absstract of: US20260244971A1

This disclosure proposes a training method for quantum machine learning and an electronic device. The training method includes: configuring a quantum circuit to output probabilities of multiple qubits, where the quantum circuit comprises multiple gates with circuit parameters; mapping the qubits to multiple model parameters of a neural network, where multiple bases are calculated based on the qubits, and the quantity of the bases is greater than or equal to the quantity of the model parameters; inputting data into the neural network and calculating a loss based on the output of the neural network; and updating the circuit parameters in the quantum circuit according to the loss.

EFFICIENTLY UPDATABLE MACHINE LEARNING

Absstract of: US20260244996A1

0000 A method of generating machine learning predictions by an efficiently updatable ensemble of machine learning models includes identifying, in response to an inference request, one or more machine learning models of the ensemble that are available for generating a prediction. An aggregated prediction is generated in response to the inference request. The aggregated prediction aggregates individual predictions generated by the one or more machine leaning models of the ensemble identified as available to generate a prediction. Responsive to determining that less than all of the ensemble of machine learning models are available, a performance guarantee based on the individual predictions is generated. The aggregated prediction is output in response to the performance guarantee satisfying a predetermined threshold.

LARGE SCALE DATA MODELING USING MACHINE LEARNING

Absstract of: US20260245105A1

0000 Examples described herein provide a computer-implemented method for large-scale data modeling using machine learning. The method includes receiving migration data from multiple countries. The method further includes integrating the migration data into a multi-modular machine learning model by: performing agent-based modeling of the migration data, performing network analysis on the migration data, and performing labor market analysis on the migration data. The method further includes performing web scraping on online resources to extract real-time or near-real-time migration-related information. The method further includes performing multi-group confirmatory factor analysis on the multi-modular machine learning model to identify underlying constructs behind how different countries shape foreign policy and migration quotas. The method further includes generating real-time suggestions for policymakers to optimize migration policies based on the multi-modular machine learning model and the real-time or near-real-time migration-related information.

MACHINE LEARNING MODEL MINIATURIZATION AND DEPLOYMENT

Absstract of: US20260244925A1

Devices, methods, and systems for automated machine learning model miniaturization and deployment are described herein. One method includes determining device specifications for a number of devices, selecting a machine learning model for the number of devices based on a function to be performed by the number of devices, selecting a miniaturization model for the machine learning model based on the function, selecting configuration settings for the miniaturization model for the number of devices, generating corresponding miniaturized machine learning models for the number of devices utilizing the selected configuration settings, and deploying the corresponding miniaturized machine learning models to each the number of devices based on the function and the device specifications associated with the number of devices.

MODEL FILE LOADING METHOD AND SYSTEM, COMPUTER DEVICE, AND STORAGE MEDIUM

Absstract of: WO2026170718A1

The present application relates to the technical field of machine learning, and discloses a model file loading method and system, a computer device, and a storage medium. The method comprises: when an inference service starts, acquiring an update request; if it is detected that the update request carries a local storage path of a model file, using the local storage path as a target storage volume of a target scheduling unit; mounting the target storage volume in an init container of the target scheduling unit, and generating a target mount item of the init container; generating response information on the basis of the target storage volume and the target mount item, wherein the response information is used for updating the target scheduling unit to establish a communication link between the updated target scheduling unit and a local directory; and sending the response information to a first slave node, so that the first slave node updates the target scheduling unit, and loads the model file on the basis of the communication link. The present application can solve problems such as high transmission delay, redundant occupation of hard disk resources, and namespace limitations during model file loading.

DISTRIBUTED TRAINING OF MACHINE LEARNING MODELS USING CONTAINER ORCHETRATION SYSTEM

Absstract of: US20260244981A1

An illustrative embodiment provides a computer-implemented method. The method comprises using a processor set to create a headless service and a number of pods for a container orchestration system. Each pod from the number of pods comprises a number of containers for performing tasks. The processor set transfers a set of training data from a cloud object storage service to the number of pods from the container orchestration system. The processor set trains a machine learning model using the set of training data. The machine learning model is trained in a distributed manner using the headless service and the number of pods for the container orchestration system, and the container orchestration system divides training task for the machine learning model into a number of portions of the training task and each pod from the number of pods performs a portion of the training task to train the machine learning model.

DYNAMIC VISUALIZATION FOR THE PERFORMANCE EVALUATION OF MACHINE LEARNING MODELS AND ANY OTHER TYPE OF PREDICTIVE MODEL

Absstract of: WO2025080778A1

A universal system and method for dynamically evaluating and visualizing the performance of any predictive model, including machine learning models. The system and method compute performance metrics based on test set data and display visual representations in real-time, allowing users to interactively explore model performance by adjusting parameters that reflect model-deployment scenarios. Key features include model-agnostic design, support for both technical and business metrics, and the ability to compare multiple models. The system and method's extensible architecture enables custom metrics and visualizations, making them scalable across various modeling use cases and industries. By providing intuitive, real-time visual feedback, embodiments of the invention empower both technical and non-technical stakeholders to gain deeper insights into model behavior, leading to more informed decisions about deployment and optimization.

TECHNIQUES FOR VERIFYING VERACITY OF MACHINE LEARNING OUTPUTS

Nº publicación: EP4792260A1 19/08/2026

Applicant:

AMKS INVEST I LLC [US]
AMKS INVESTMENTS I LLC

Absstract of: US11966704B1

The techniques described herein relate to techniques for verifying a veracity of machine learning outputs. An example method includes receiving a first output generated by a first model responsive to a first input, the first output comprising one or more verifiable statements in text, verifying, using a second model and first reference data stored in at least one first datastore, the one or more verifiable statements to produce first verification results indicating which of them has been verified, when it is determined that at least one of them remains unverified based on the first verification results, identifying, using at least one of the first or second models, at least one second datastore having second reference data attesting to veracity of the first output; and verifying, using the second model and the second reference data, the at least one unverified statement to produce second verification results to be provided as output.