Patents.

Computer-based platforms and systems configured for event processing with a specialized domain-specific programming language and methods of use thereof

Deep learning-based eddy current correction

Systems and methods for providing improved eddy current correction (ECC) in medical imaging environments. One or more of the embodiments disclosed herein provide a deep learning-based convolutional neural network (CNN) model trained to automatically generate an ECC mask which may be composited with two-dimensional (2D) scan slices or four-dimensional (4D) scan slices and made viewable through, for example, a web application, and made manipulable through a user interface thereof.

Systems and Methods for Configuring Artificial Intelligence Agents

Systems and methods for configuring artificial intelligence agents to carry out user-specified tasks against software interfaces.

Systems and methods for vision-language distribution alignment

Embodiments described herein a CROss-Modal Distribution Alignment (CROMDA) model for vision-language pretraining, which can be used for retrieval downstream tasks. In the CROMDA mode, global cross-modal representations are aligned on each unimodality. Specifically, a uni-modal global similarity between an image/text and the image/text feature queue are computed. A softmax-normalized distribution is then generated based on the computed similarity. The distribution thus takes advantage of property of the global structure of the queue. CROMDA then aligns the two distributions and learns a modal invariant global representation. In this way, CROMDA is able to obtain invariant property in each modality, where images with similar text representations should be similar and vice versa.

Training conversational computing interfaces with traced programs

Asynchronous distributed data flow for machine learning workloads

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for distributing machine learning workloads, e.g., computations for training a neural network or computing an inference using a neural network, across multiple hardware accelerators. One of the systems comprises a plurality of accelerator islands, each hardware accelerator island comprising a respective plurality of hardware devices that include a plurality of hardware accelerators and a corresponding host for each of the plurality of hardware accelerators; and a respective scheduler for each of the accelerator islands that is configured to schedule workloads across the plurality of accelerators and corresponding hosts in the accelerator island, wherein the system is configured to: receive data representing a machine learning workload; and assign a respective portion of the machine learning workload to each of the plurality of accelerator islands for scheduling by the respective scheduler for the accelerator island.

Rights management regarding user data associated with data lifecycle discovery platform

Generation of Agentic Trajectories for Training Artificial Intelligence Agents

A system for generating training data to train agents to automate tasks otherwise done by users, using an intermediary between an interface and a user that intercepts user-actuated actions to build agentic trajectories.

Universal logout and single logout techniques

A method of identity management is described. The method may include receiving a request to terminate two or more sessions between a user of an identity management system and two or more respective applications that are accessible via the identity management system. The request may include information associated with the user, information associated with the two or more sessions, or both. The method may further include receiving, via an endpoint associated with a tenant of the identity management system, a set of application programming interface (API) credentials that are usable to communicate with the two or more applications via two or more respective APIs. The method further includes transmitting, via the two or more APIs, respective API calls to terminate the two or more sessions between the user and the two or more applications in accordance with a universal logout (ULO) operation or a single logout (SLO) operation.

Distinct value estimation for query planning

Information retrieval system using a hierarchical corpus encoder

A dense encoder is adapted as a hierarchical corpus encoder in an information retrieval system to use negative samples from sibling nodes in a hierarchical tree of vector embeddings for documents in a corpus. Both the encoder and hierarchical tree are co-trained using a loss function that takes the document hierarchy into account. The hierarchical corpus encoder may be used in both supervised training cases where query-document relevance judgments are present and in zero-shot cases where a query dataset is absent.

Rewriting sentences for non-inclusion of words of interest with contextual accuracy

Rewriting sentences for non-inclusion of words of interest with contextual accuracy — an NLP method for detecting words of interest in text and generating replacement phrasing that preserves the original meaning. (Priority 2023-09-30; co-assigned IBM Deutschland GmbH.)

Domain-Aware Neurosymbolic Agents For Improving Problem-Solving Accuracy And Consistency

Evaluating effects of an artificial intelligence model on enterprise performance objectives

Systems, computer-implemented methods, and/or computer program products facilitating a process to monitor and evaluate the effects of an artificial intelligence (AI) model on enterprise performance metrics are provided. According to an embodiment, a computer implemented method can comprise determining a technical issue of candidate technical issues associated with an artificial intelligence model that correlates to a change associated with a performance metric, wherein the determination is based on using a first data model that defines first relationships between the key performance metrics and candidate technical issues and second relationships between the candidate technical issues and candidate solutions. The method further comprises determining a solution for the technical issue using the data model and recommending or automatically implementing the solution. The method further provides for updating/refining the data model over time using continuous learning based on evaluating whether and how implemented solutions impact the relevant performance metrics.

Systems and methods for field extraction from unlabeled data

A field extraction system that does not require field-level annotations for training is provided. Specifically, the training process is bootstrapped by mining pseudo-labels from unlabeled forms using simple rules. Then, a transformer-based structure is used to model interactions between text tokens in the input form and predict a field tag for each token accordingly. The pseudo-labels are used to supervise the transformer training. As the pseudo-labels are noisy, a refinement module that contains a sequence of branches is used to refine the pseudo-labels. Each of the refinement branches conducts field tagging and generates refined labels. At each stage, a branch is optimized by the labels ensembled from all previous branches to reduce label noise.

Event extraction from documents with co-reference

Methods, systems, and computer-readable media for event extraction from documents with co-reference are disclosed. An event extraction service identifies one or more trigger groups in a document comprising text.

Systems and methods for generating source-agnostic trajectories

System and method for application security profiling

A system and method for application security profiling that includes extracting a code property graph from at least a subset of a code base; generating a code profile from the code property graph, wherein generating the code profile occurs prior to a compilation of the code base; and applying the code profile, comprising of identifying sections of interest within the code base.

Vehicle control system and method

A vehicle control system for controlling a vehicle, including a set of task nodes, wherein each of the set of task nodes compares a failure criterion to input to the task node to generate a failure flag and provides output to at least one other task node of the set of task nodes; an aggregator node directly communicatively coupled to each of the set of task nodes, and wherein the aggregator node: receives a set of failure flags from the set of task nodes, determines a failure tier based on the set of failure flags, and generates a failure tier message defining the failure tier; and, a behavior planning node communicatively coupled to the aggregator node, wherein the behavior planning node: receives the failure tier message, and in response to the failure tier defined by the failure tier message, generates vehicle control instructions.

Food description processing methods and apparatuses

Health tracking system with verification of nutrition information

Utilizing machine learning to generate an application program

An input specifying a schematic of user interface components of an application program is received. A first group of one or more machine learning models is used to automatically identify the user interface components and associated properties specified in the input. Based on the identified user interface components and the associated properties, a second group of one or more machine learning models is used to automatically generate program code implementing the application program including the user interface components.

Time-based ensemble machine learning model

An input dataset is sorted into a first version of data and a second version of data. The first version of data is associated with a first period of time and the second version of data is associated with a second period of time. The second period of time is a shorter period of time than the first period of time. A first set of one or more machine learning models is generated based on the first version of data. A second set of one or more machine learning models is generated based on the second version of data. The first set of one or more machine learning models and the second set of one or more machine learning models are combined to generate an ensemble model. A prediction based on the ensemble model is outputted. The prediction indicates abnormal behavior associated with the input dataset.

Automated configuration of augmented and virtual reality avatars for user specific behaviors

Identifying and authenticating users based on passive factors determined from sensor data

Structured Cluster Execution for Data Streams

A system for executing a streaming query includes an interface and a processor. The interface is configured to receive a logical query plan. The processor is configured to determine a physical query plan based at least in part on the logical query plan. The physical query plan comprises an ordered set of operators. Each operator of the ordered set of operators comprises an operator input mode and an operator output mode. The processor is further configured to execute the physical query plan using the operator input mode and the operator output mode for each operator of the query.

Evolved machine learning models

A plurality of initial machine learning models are determined based on a plurality of original features. The plurality of initial machine learning models are filtered by selecting a subset of the initial machine learning models as one or more surviving machine learning models. One or more evolved machine learning models are generated. At least one of the evolved machine learning models is based at least in part on one or more new features, which are based at least in part on a transformation of at least one of features of the one or more surviving machine learning models. Corresponding validation scores associated with the one or more evolved machine learning models and corresponding validation scores associated with the one or more surviving machine learning models are compared.

Domain Specific Language for Generation of Recurrent Neural Network Architectures

A system automatically generates recurrent neural network (RNN) architectures for performing specific tasks, for example, machine translation. The system represents RNN architectures using a domain specific language (DSL). The system generates candidate RNN architectures. The system predicts performances of the generated candidate RNN architectures, for example, using a neural network. The system filters the candidate RNN architectures based on their predicted performance. The system generates code for selected a candidate architectures. The generated code represents an RNN that is configured to perform the specific task. The system executes the generated code, for example, to evaluate an RNN or to use the RNN in an application.

Key-value storage using a skip list

Distributed system of computer architectures

Provided herein are methods for communicating data between nodes in a computer system and distributed systems of computer architectures. The methods include generating a hash based upon a data set, communicating the hash to one or more nodes, comparing the communicated hash to stored hashes at the nodes.

Determining repair instructions in response to natural language queries

Techniques for determining repair instructions in response to natural language queries. (abstract excerpt from Google Patents)

System and method for associating user-entered text to database entries

Systems and methods for detecting photoplethysmographic device usage

Systems and methods for determining if a wearable photoplethysmography device is correctly positioned in operating to measure medical signs of a user by using a classifier to determine if a signal is valid or invalid.

Search space exploration for deep learning

Aspects of the invention include systems and methods to obtain meta features of a dataset for training in a deep learning application. A method includes selecting an initial search space that defines a type of deep learning architecture representation that specifies hyperparameters for two or more neural network architectures. The method also includes applying a search strategy to the initial search space. One of the two or more neural network architectures are selected based on a result of an evaluation according to the search strategy. A new search space is generated with new hyperparameters using an evolutionary algorithm and a mutation type that defines one or more changes in the hyperparameters specified by the initial search space, and, based on the mutation type, the new hyperparameters are applied to the one of the two or more neural networks or the search strategy is applied to the new search space.

Dummy class framework for continual supervised learning applications

System and Method for Securing a Mobile Services Account

The disclosed technology relates to a system for securing account information using blockchain technology. The system links the configuration of a user's account to a blockchain wallet. Changes to the account configuration are pre-approved in a blockchain transaction signed by the owner's wallet private-key. The account configuration is hashed and stored on the blockchain. The system includes a monitoring component that periodically checks the status of account details and compares the hash on the blockchain with the hash of the current network observed configuration. If the hashes do not match, the system can disable the account.

Goal-based dynamic modifications to user interface content

Extracting custom content from network packets

Interactive gaming system

A system for customizing an end user gaming experience is disclosed herein. The system includes a processor. The processor is configured to create a campaign by receiving a campaign creation request from an end user device, the campaign creation request comprising one or more parameters of a campaign associated with the campaign request, creating the campaign based on the one or more parameters associated with the campaign request, and personalizing one or more offers for an end user, each of the one or more offers comprising a game having one or more hurdles, wherein completion of each of the one or more hurdles leads to a reward. The processor is further configured to map a subset of the one or more personalized offers to the end user. The processor is further configured to monitor progress of the campaign.

Systems and methods for non-invasive blood pressure measurement

Systems and methods for non-invasive blood pressure measurement are disclosed. In some embodiments, a system comprises a wearable member configured to generate first and second signals, and a blood pressure calculation system with pre-processing, wave selection, feature extraction, and blood pressure processing modules.

Using machine learning models to analyze contractual terms and clauses in a legal contract, to recommend edits, and to make changes to a workflow

Aspects discussed herein may relate to using machine learning models as part of methods and techniques for ingesting, creating, storing, editing, and managing a document. The document may be a legal contract that includes one or more clauses. Among other things, one or more machine learning models may be configured to recognize clauses and/or classifications, or types, of clauses. For example, the one or more generative language models may be used to generate one or more recommended edits to a clause, generate one or more suggested clauses that are missing from the contract, and/or generate one or more suggested locations where a clause may be inserted into or moved within the contract.

Adaptively adjusting influence in federated learning model updates

Techniques for improved federated learning are provided. One or more queries are issued to a plurality of participants in a federated learning system, and one or more replies are received from the plurality of participants. A first aggregated model is generated based on the one or more relies and a first influence vector. Upon determining that a predefined criterion is satisfied, a second influence vector modifying a weight of a first participant of the plurality of participants is generated. A second aggregated model is generated based on the one or more replies and the second influence vector.

Data validation for automatic model building and release

URL rewriting

Determining an effect of a message on a personal brand based on future goals

Provided are techniques for determining an effect of a message on a personal brand based on future goals. A message is received from an entity for a social media platform. Additional messages related to the message are identified. A reaction sentiment trend is determined for each topic of the additional messages for a period of time. An effect the message has on a personal brand of the entity in future is predicted based on future goals of the entity and based on the reaction sentiment trend for each topic. An indication of whether the message is aligned with the future goals based on the predicted effect is provided. In response to the message being aligned with the future goals, the message is posted to the social media platform. In response to the message not being aligned with the future goals, one or more suggestions to modify the message are provided.

Determining repair status information using unstructured textual repair data in one or more natural languages

Techniques are described for performing automated operations related to identifying and using repair and maintenance status information, such as summarizing and encoding such information for one or more repair areas or other target domains. (abstract excerpt from Google Patents)

Predicting medical device failure based on operational log data

Distributed hyperparameter tuning and load balancing for mathematical models

A device generates a first set of combinations of hyperparameter values associated with training a mathematical model (including a machine learning model) and distributes the tuning work with load balancing.

Image processing neural networks with separable convolutional layers

Methods and systems for calibration of multiple lidar devices with non-overlapping fields of view

A method and system for calibration of a first lidar device and a second lidar device with non-overlapping fields of view. Pairs of vehicle positions along a predetermined path pattern are determined such that the field of view of the first lidar at one position overlaps the field of view of the second lidar at another position, and the calibration is computed from those pairs.