Here is generation of a local explanation of a machine learning (ML) inference, and this generation is accelerated by accurate estimation of local feature value importances from a multioutput regression by an attribution metamodel. Into a feature vector, a computer stores values for respective features and stores an inference, from the feature values, by an ML model. A multioutput regression by an attribution metamodel includes inferentially generating, from the feature vector, local value importances respectively for the feature values. Based on the local value importances, a local explanation of the inference is generated and displayed. For accelerated training with early stopping, an adaptively selected training corpus involves incrementally adding additional datapoints to the training corpus until the attribution metamodel accurately learns.
A method implements generating points-to relations using language models. The method involves executing a preprocessing session using one or more language models to process source code using one or more preprocessing prompts to generate metadata from the source code. The method further involves executing an analysis session using the one or more language models to process the metadata using an analysis prompt to generate a points-to report for the source code from the metadata. The method further involves executing an evaluation session using the one or more language models to process the points-to report and normalized code from the metadata using an evaluation prompt to generate a revised points-to report. The revised points-to report includes a confidence score for an entry in the points-to report.
Techniques discussed herein relate to generating synthetic training data with which a multi-speaker Text-To-Speech model may be trained. Text and an audio sample of a speaker may be provided as input to a voice generation system (VGS). The VGS may generate, based on audio features of the audio sample, machine-generated audio comprising a synthetic voice providing the text as spoken words. A synthetic training example including the machine-generated audio may be generated and combined with synthetic training data examples corresponding to a second speaker (and/or with training data examples comprising recordings of a second speaker). A multi-speaker Text-To-Speech model may be trained with a training set comprising the synthetic training data examples to generate subsequent audio that provides spoken words of input text in a synthetic voice that is generated to replicate features associated with one of a plurality of speakers comprising the first speaker and the second speaker.
A system that trains a machine learning model to determine capture levels for target computations is disclosed. The system utilizes training data encompassing attribute sets and information levels for various computation types. For a form field value computation, the system determines associated attributes. The trained model processes these attributes to establish an appropriate information storage level. Based on this level, the system selects a relevant subset of information related to the computation or its result. The system then stores this selected subset in association with the computed value. The system uses feedback to retrain the model to enhance its performance.
Techniques for evaluating a language model are disclosed herein. One or more agents and/or one or more language models are used to generate responses to queries in conversations. For example, during chat session conversations, the agent and/or model used to generate responses to queries in the conversation is sometimes changed. Various metrics are collected for the conversations for which a prompt, agent or model change occurs and/or the conversations for which a change to the prompt, agent and/or model does not occur. Metrics such as success rate, escalation rate, query repetition, direct feedback, or other feedback is collected and/or aggregated for groups of users and/or topics of conversations. One or more models used to generate responses in the conversations are evaluated and/or selected based on the metrics.
Systems, methods, and computer-readable media are provided for intelligent code integration, deployment, and analysis. An example intelligent code analysis system is configured to access and maintain an issue knowledge base storing information about code issues and/or issue resolutions associated with one or more software pipelines managed by the system. The system may process a natural language request about one or more code issues, issue resolutions, and/or sets of code. The system may additionally or alternatively analyze code issues to determine sets of code that should be newly associated with the issues, and/or analyze sets of code to determine issues that should be newly associated with the sets of code. The system generates a prompt that includes content from the natural language request, information about code issues, and/or information about sets of code. The system prompts a large language model (LLM) with the prompt and provides a response for the natural language request based on a result of execution of the prompt by the LLM.
A computer comprising one or more processors and memory may implement multiple threads that perform synchronization operations using a semaphore comprising an allocation field and a grant field. Upon entry to a semaphore wait operation, a thread allocates a ticket by atomically copying a ticket value contained in the allocation field and incrementing the allocation field. The thread computes a difference between the allocated ticket to the grant field representing a number of waiting threads. If the number is above the threshold, the thread enters a long term wait operation comprising determining a location for long term wait value and waiting on changes to that value. If the number is below the threshold or the long term wait operation is complete, the thread waits for the grant value to equal the ticket to complete waiting on the semaphore.
Disclosed is a system for positioning fiber and electronics cables within a server room, which includes a wire-pulley system which includes a wire operably coupled to a first pulley wheel and a second pulley wheel. The wire is looped around the first and second pulley wheel such that a point in the wire is laterally movable between the first and second pulley wheels when the first and second pulley wheels are rotated. The system for positioning cables includes a cable carrier which is removably coupled to the point in the wire. The cable carrier includes a first panel, and a second panel hingedly coupled to the first panel. The first panel and second panel each include a plurality of receiving slots, where the receiving slots are configured to removably receive a distal end of various cables.
H02G 1/04 - Methods or apparatus specially adapted for installing, maintaining, repairing, or dismantling electric cables or lines for overhead lines or cables for mounting or stretching
B65G 17/20 - Conveyors having an endless traction element, e.g. a chain, transmitting movement to a continuous or substantially-continuous load-carrying surface or to a series of individual load-carriersEndless-chain conveyors in which the chains form the load-carrying surface comprising load-carriers suspended from overhead traction chains
9.
SECURITY ENGINE FOR MONITORING EDGE NODES AND ASSOCIATED DEVICES WITHIN A DISTRIBUTED NETWORK
Various embodiments of the present technology generally relate to systems and methods for providing a security engine. In an example, the security engine determines a plurality of devices sharing a common connection to a network and groups these devices into a first group based on the common connection. The security engine then determines group attributes for the plurality of devices within the first group. The group attributes include at least one attribute for each device of the plurality of devices. Based on the group attributes, the security engine generates a group identifier for the first group. Using the group identifier, the security engine then monitors the devices, in particular the common connection, to detect any attribute changes. If the security engine detects an attribute change, the security engine determines whether the attribute change indicates potential malicious activity, and if it does, generates an alert based on the attribute change.
Techniques for populating the fields of a form are disclosed. A machine learning model may be trained to predict associations between fields. The trained machine learning model may be applied to a plurality of forms to predict an association between a first field in a first form type and a second field in a second form type. Upon receiving a value for a first field in a first form of a first form type and based on the predicted association, the system may populate a second field of a second form of the second form type based on the value.
A computer generates and iteratively improves operational logic by incremental refinement of a linguistic prompt to increase its accuracy or by full replacement of the linguistic prompt by a somewhat unrelated linguistic prompt. From a natural language (NL) question about data, a problem summarization large language model (LLM) generates an NL problem summary that specifies multiple semantic requirements. From the semantic requirements, a prover LLM generates a test case that can detect an error in an answer for the NL question. From the semantic requirements, a strategy LLM generates multiple NL solution summaries that each describes a respective distinct solution. An initially best solution summary is iteratively reused to generate and refine solution logic. Excessive iterating may cause switching to a different solution summary for solution logic generation. The generated test cases ensure that logic generation continues until an answer for the NL question is obtained that is provably correct.
For a new batch of datapoints during a detected data drift, without decreasing accuracy nor increasing volatility of accuracy, noniterative selection of a new hyperparameters configuration accelerates generating a retrained machine learning model. For each batch in a sequence of recent batches of datapoints that includes a last batch of datapoints, a computer stores: the batch, a hyperparameter value for an unsupervised model, and an accuracy measurement, for the unsupervised model, that is based on both of the hyperparameter value and the batch. With a retraining corpus that contains the new batch and the batches in the sequence of recent batches, a new hyperparameter value is predicted. Based on the retraining corpus and the new hyperparameter value, a new accuracy measurement is estimated. In response to detecting that the new accuracy measurement exceeds an accuracy threshold, retraining the unsupervised model with the retraining corpus is unconventionally accelerated.
Techniques for predicting render times are disclosed. A system accesses auxiliary rendered outputs that were generated from a coarse render pass of a 3D scene description using a set of hardware resources. The system accesses system performance metrics of the set of hardware resources associated with the coarse render pass. The system encodes visual feature representations from the auxiliary rendered outputs using a convolutional neural network. The system also encodes the system performance metrics and a set of rendering configuration parameters associated with a target rendering operation. The system generates, using a multi-modal fusion model, a fused representation of the encoded visual feature representations, the encoded system performance metrics, and the encoded first set of rendering configuration parameters, using an attention-based mechanism that weights contributions of the inputs to the model. The system predicts a render time for the target rendering operation based on the fused representation.
Techniques for drilling back to an original audio clip in virtual assistant initiated lists and reminders are disclosed. The system may receive audio input comprising a first request. Based on the first request, the system may schedule an action to be performed by the virtual assistant platform. The system stores at least a portion of the audio input and a mapping between the action and at least the portion of the audio input. The system performs the action. Subsequent to performing the action, the system receives a second request for audio playback of the first request corresponding to the action. The system retrieves at least the portion of the audio input based on the mapping between the action and at least the portion of the audio input, and plays at least the portion of the audio input comprising the first request.
G10L 15/06 - Creation of reference templatesTraining of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
G10L 15/22 - Procedures used during a speech recognition process, e.g. man-machine dialog
15.
CHECKING FOR POTENTIAL SOURCE CODE HAVING AN ISSUE BY AUGMENTING A NATURAL LANGUAGE PROMPT WITH SOURCE-CODE-DESCRIBING CONTENT THAT IS SIMILAR TO THE ISSUE
Systems, methods, and computer-readable media are provided for intelligent code integration, deployment, and analysis. An example intelligent code analysis system is configured to access and maintain an issue knowledge base storing information about code issues and/or issue resolutions associated with one or more software pipelines managed by the system. The system may process a natural language request about one or more code issues, issue resolutions, and/or sets of code. The system may additionally or alternatively analyze code issues to determine sets of code that should be newly associated with the issues, and/or analyze sets of code to determine issues that should be newly associated with the sets of code. The system generates a prompt that includes content from the natural language request, information about code issues, and/or information about sets of code. The system prompts a large language model (LLM) with the prompt and provides a response for the natural language request based on a result of execution of the prompt by the LLM.
Techniques are disclosed herein for automatically curating preference alignment training (PAT) data and subsequently using the PAT data to train a machine-learning model used to perform natural language-to-query language tasks. A first PAT dataset is generated using a first trained LLM in an execution-based data generation technique. A second PAT dataset is generated using a second trained LLM in a LLM-based data generation technique. The second PAT dataset is populated with synthetic data points generated by the second trained LLM. Each data point of first PAT dataset and the second PAT dataset comprises a natural language query and an executable and non-executable query language statement. The first PAT dataset and the second PAT dataset can be combined and filtered to generate a PAT dataset that is optimized for training a machine-learning model to overcome one or more weaknesses exhibited when generating query language statements responsive to natural language questions.
In one embodiment, a software is operable when executed to receive a first user request from a first user to implement a change in a security policy, detect a dilution in security posture based on the implemented change in the security policy, request a first acknowledgement from the first user that the implemented change constitutes the dilution in the security posture responsive to detecting the dilution in the security posture, record the first acknowledgement from the first user that the implemented change constitutes the dilution in the security posture in a database, and determine whether a security breach occurred within a time period when the dilution in the security posture existed using the database.
Techniques for dynamically loading and displaying hierarchical data are disclosed. A system selectively loads node data from a datastore storing a node hierarchy by determining if a particular node meets a display criterion and corresponds to an expanded node state. If a system determines that the node meets the display criterion but corresponds to a collapsed node state, the system refrains from loading the node data from the data store to a local cache. If the system determines the node corresponds to the expanded node state but does not meet the display criterion, the system does not load the node data from the datastore to the local cache.
In one embodiment, a software is operable when executed to identify a log indicating executions of commands of a secure shell (SSH) session, wherein a user accesses an instance through the SSH session, identify a generic prompt that specifies a generic set of commands of interest, identify a customized prompt that specifies a customized set of commands of interest, the customized set of commands of interest being specific to an attribute associated with one or more of the user, the instance, or the SSH session, input the generic prompt, the customized prompt, and the log to a generative AI model, and generate, using the generative AI model, a risk assessment for the SSH session based on one or more executions of the generic set of commands of interest and one or more executions of the customized set of commands of interest from the executions of the commands in the log.
G06F 21/57 - Certifying or maintaining trusted computer platforms, e.g. secure boots or power-downs, version controls, system software checks, secure updates or assessing vulnerabilities
20.
CORRECTION BASED PREFERENCE ALIGNMENT FINE-TUNING FOR LOW RESOURCE PROGRAMMING LANGUAGES
Techniques are disclosed herein for generating fine-tuning examples of low-resource programming languages and enhancing code generation capabilities by using the examples in PAT of generative models such as LLMs. The techniques focus on the use of generative models that possess version-specific knowledge of low-resource programming language (e.g., SuiteScript 2.1) to correct errors in synthesized fine-tuning data. In each preference pair, the “chosen” example is the synthesized fine-tuning data after correction, and the “rejected” example is the fine-tuning data before correction. By employing this approach in a multi-stage pipeline, refined preference pairs can be generated that are then used for preference alignment fine-tuning of generative models.
A cloud-computing service (e.g., a “Puffin Service”) is described. The service may maintain service and skill catalogs corresponding to various services to be deployed to a region (e.g., during a region build). The service may host numerous user interfaces with which various service and skill metadata may be provided. In some embodiments, such data may include one or more dependencies between skills. The data managed by the cloud-computing service may be utilized to build a dependency graph. Navigation of the dependency graph may be performed via one or more user interfaces hosted by the cloud-computing service. An orchestration service (e.g., a Multi-Flock Orchestrator) may manage bootstrapping efforts for any suitable number of services during a region build based at least in part on dependencies between skills.
Systems, methods, and machine-readable media to migrate data from source databases to target databases are disclosed. Data may be received, relating to the source databases and the target databases. For each source database, a migration assessment may be generated based on analyzing the data, and a migration method may be selected. A migration plan that specifies a parallel migration of a set of databases to the target databases may be created, with a first migration method to migrate a first subset of the set of databases and a second migration method to migrate a second subset of the set of databases. The parallel migration may be executed according to the migration plan may be caused so that the first subset of the set of databases is migrated with the first migration method while the second subset of the set of databases is migrated with the second migration method.
G06F 16/27 - Replication, distribution or synchronisation of data between databases or within a distributed database systemDistributed database system architectures therefor
23.
Using Data Submitted For A Field To Populate A Different, Associated Field
Techniques for populating the fields of a form are disclosed. A machine learning model may be trained to predict associations between fields. The trained machine learning model may be applied to a plurality of forms to predict an association between a first field in a first form type and a second field in a second form type. Upon receiving a value for a first field in a first form of a first form type and based on the predicted association, the system may populate a second field of a second form of the second form type based on the value.
Techniques are disclosed for avoiding termination indeterminism associated with a phase of concurrent operations. For instance, during a concurrent marking phase of a garbage collection cycle, the system generates a first data structure to record reachability information of runtime objects in a memory area that are strongly reachable, and the system generates a second data structure to record reachability information of runtime objects that are reachable but not strongly reachable. In particular, the second data structure records reachability information of runtime objects that could become strongly reachable as a result of a potential resurrection. If the potential resurrection occurs during a termination procedure for the concurrent marking phase, the system takes note of the resurrection and proceeds with the termination procedure as normal. The system uses the second data structure to update the reachability information recorded by the first data structure after the termination procedure is completed.
Techniques are disclosed for managing the processing of pointers during a concurrent phase of operations using a monotonic state machine. The monotonic state machine defines multiple distinct state designations. During the phase of concurrent operations, the state of any given pointer may be described by a state designation defined by the monotonic state machine. The state designations defined by the state machine have stacking requirements. For example, a third state designation corresponding to a greater level of accessibility necessarily implies a second state designation corresponding to a lesser level of accessibility. The system models changes to a pointer's state during the phase of concurrent operations as monotonic transitions between state designations. If a pointer is transitioned from a former state designation to a latter state designation during the phase of concurrent operations, the pointer is not reverted backwards to the former state designation during the phase of concurrent operations.
Techniques are disclosed for performing a garbage collection process that processes reference objects in multiple generations of a memory area while a program instance is concurrently manipulating runtime objects in the memory area. During one phase of concurrent operations, the system applies one rule set for processing reference objects in one generation of the memory area. During another phase of concurrent operations, the system applies another rule set for processing reference objects residing in another generation of the memory area. To enable the performance of the one phase of concurrent operations and the other phase of concurrent operations at the same time, the system tracks the accessibility of a given runtime object with respect to both of the one generation and the other generation. The system tracks the given runtime object's accessibility with respect to both generations by embedding metadata into a pointer that refers to the given runtime object.
Generative natural language processing (NLP) and source logic refactoring by large language models (LLMs) are deferred until interactively-initiated deployment of a Python notebook. The deployment is interactively initiated from a Python kernel in a computer. An LLM inferentially detects many extraneous cells in an original sequence of cells in an original Python notebook. An LLM inferentially generates a modified copy of the original Python notebook. For each extraneous cell, the extraneous cell is excluded from representation in the modified notebook based on a cell type of the extraneous cell. New subroutines are generated and included in the modified notebook. Differential tests and natural language for reference documentation are inferentially generated for the modified notebook.
Various embodiments of the present technology generally relate to systems and methods for providing a network function (NF) availability engine. In an example, a first NF availability executing on a first NF instance registers a minimum availability threshold for a first NF set associated with the first NF instance with a Network Repository Function (NRF), where the minimum availability threshold defines a minimum number of NF instances required to maintain a fault tolerance level of the first NF set. The NRF then receives a discovery request containing discovery parameters associated with a first session from a consumer NF. An NF availability engine executing on the NRF determines whether the minimum availability threshold for the first NF set is satisfied and if so, generates a listing of discoverable NF profiles including a first NF profile associated with the first NF instance based on the minimum availability threshold being satisfied.
H04L 41/0681 - Configuration of triggering conditions
H04L 43/0817 - Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability by checking functioning
H04L 45/00 - Routing or path finding of packets in data switching networks
H04L 67/51 - Discovery or management thereof, e.g. service location protocol [SLP] or web services
29.
OPTIMIZATION FRAMEWORK FOR ENHANCING DOMAIN ADAPTATION OF GRAPHRAG VIA ENTITY-RELATION OPTIMIZATION
Techniques are disclosed herein for summarizing complex and lengthy documents in various domains. In one aspect, a method includes: extracting, by a generative model, entities and their relationships from source content based on entity types and examples of entity relationships, generating, by a generative model, a network graph based on the extracted entities and their relationships, and generating, by a generative model, one or more summaries of the source content based on the source content and the network graph. The summaries are then evaluated using an optimizer generative model, optimizer prompt, and objective function against ground truth summaries. One or more recommendations are generated based on the evaluation to improve one or more prompts used for extracting the entities and their relationships from source content to better align entities and their relationships with specified objectives.
Techniques are provided for enhancing the capabilities of generative models in tasks such as causal reasoning by leveraging feedback from advanced reasoning models. The disclosed approach features an automated, iterative alignment process in which a generative model generates responses to training examples using an initial prompt. An optimizer model evaluates each response against a reference answer based on a specified objective function, producing alignment instructions to improve the generative model's outputs. These alignment instructions are used to refine the prompt, directing the generative model toward closer agreement with reference answers and better performance on desired metrics. The process begins with a default prompt and, through successive iterations, employs optimized prompts derived from the alignment instructions. Upon completion, the system outputs the final alignment instructions, facilitating improved task performance through model alignment and prompt optimization.
Techniques are disclosed herein for implementing planning mechanisms for agentic artificial intelligence (AI) systems. A query concerning a person and a request for a context identifier is received from a client device. A data set describing characteristics for a set of persons is accessed. The context identifier is generated and transferred to the client device. An event message including an event associated with the query and the context identifier is received from the client device. A determination is made that a characteristic for the person is included in characteristics and an input for a generative machine learning model is generated. An execution plan for generating a response to the query is obtained using the generative machine learning model and the input and a response to the query is generated by executing the execution plan.
Techniques disclosed herein pertain to agentic artificial intelligence (AI) systems, and, more specifically, to safety mechanisms for agentic AI systems. A query is accessed, and, in response to accessing the query, a planning process to generate an execution plan for responding to the query and a safety analysis process to determine a safety classification for the query based on a set of policies can be initiated. A determination can be made as to whether to authorize execution of the execution plan based on the safety classification from the safety analysis process. In response to determining to authorize execution of the execution plan, the execution plan can be executed and a substantive response to the query can be generated based on the execution. In response to determining not to authorize execution of the execution plan, execution of the execution plan can be blocked and a fallback response can be generated.
Techniques are described herein for advanced persistent threat detection. An example method can include a device receiving a message identifying an instruction loaded onto an instruction cache of a secure processer. The device can transmit a control instruction to configure a kernel image to collect a metric over a first time interval, the metric being generated based at least in part on the secure processor executing the instruction during the first time interval. The device can receive the first metric from the kernel image, the metric being indicative of a transition of the secure processor from a non-secure state to a secure state. The device can determine whether the secure processor is undergoing a computing attack based on the metric. The device can transmit the determination of whether the secure processor is undergoing a computing attack to a sender of the message.
In some implementations, techniques described herein may include identifying text in a visually rich document and determining a sequence for the identified text. The techniques may include selecting a language model based at least in part on the identified text and the determined sequence. Moreover, the techniques may include assigning each word of the identified text to a respective token to generate textual features corresponding to the identified text. The techniques may include extracting visual features corresponding to the identified text. The techniques may include determining positional features for each word of the identified text. The techniques may include generating a graph representing the visually rich document, each node in the graph representing each of the visual features, textual features, and positional features of a respective word of the identified text. The techniques may include training a classifier on the graph to classify each respective word of the identified text.
The present disclosure provides dynamic routing for data flows to a customer network hosted in the cloud. A plurality of compute instances may share a common virtual IP address. Each of the plurality of compute instances may advertise information to a respective network virtualization device (NVD). The information may include the IP address, cost, and/or active/standby status of the compute instance. The NVD may then provide the information to the control plane of a virtual cloud network (VCN), which may aggregate the information from the plurality of compute instances and generate a forwarding table, which may be sent to the NVDs. These techniques may allow a customer to automatically remove a compute instance whose service host has failed. These techniques may also allow a customer to add compute instances and to route data flows according to an active-standby operation, an equal cost active-active operation, or an unequal cost active-active operation.
09 - Scientific and electric apparatus and instruments
35 - Advertising and business services
41 - Education, entertainment, sporting and cultural services
42 - Scientific, technological and industrial services, research and design
Goods & Services
Computer hardware. Arranging, organizing, and conducting trade shows,
exhibitions, and business networking events in the fields of
computer software, computer hardware, computer peripherals,
cloud-based services, cloud computing, cloud-based
platforms, cloud services, computer and cloud applications,
business automation, computer networking, artificial
intelligence, technology planning, business management, and
product and technology demonstrations. Arranging and conducting educational conferences, seminars,
speeches, and entertainment, namely, in-person and virtual
presentations, all in the fields of computer software,
computer hardware, computer peripherals, cloud-based
services, cloud-based platforms, cloud services, computer
and cloud applications, business automation, computer
networking, artificial intelligence, technology planning,
business management, and product and technology
demonstrations; arranging and conducting educational
conferences featuring business leaders, developers, IT
management, and business end-users of enterprise technology,
cloud computing, cloud platforms, cloud services, computer
and cloud applications, and business automation technology;
arranging and conducting in-person and virtual business
conferences featuring business leaders, developers, IT
management, and business end-users of enterprise technology,
cloud computing, cloud platforms, cloud services, computer
and cloud applications, and business automation technology. Software as a service (SaaS) featuring software for
creating, managing, and deploying web applications, data
storage and backup, database management, virtualization,
networking, collaboration, access, support, cloud computing,
data sharing, data security, artificial intelligence,
administration and management of computer software and
hardware, and distribution and transmission of data and
information; platform as a service (PAAS) featuring computer
software and hardware platforms for creating, managing, and
deploying web applications, data storage and backup,
database management, virtualization, networking,
collaboration, access, support, cloud computing, data
sharing, data security, artificial intelligence,
administration and management of computer software and
hardware, and distribution and transmission of data and
information; infrastructure as a service (IAAS) featuring
computer software and hardware platforms for creating,
managing, and deploying web applications, data storage and
backup, database management, virtualization, networking,
collaboration, access, support, cloud computing, data
sharing, data security, artificial intelligence,
administration and management of computer software and
hardware, and distribution and transmission of data and
information; databases as a service (DBaaS), namely, cloud
hosting of computer databases for others; platform as a
service (PaaS) featuring customizable software platform for
collecting, analyzing, integrating, transforming, and
providing access to data and information; providing online
non-downloadable customizable software platform for
collecting, analyzing, integrating, transforming, and
providing access to data and information; software as a
service (SaaS) services featuring software to collect,
manage, analyze, retrieve, monitor, maintain, report on,
structure, model, forecast, present and display data and
information in the fields of marketing, sales, retail
services, customer service, hospitality, contracts, human
resources, clinical research, health care, health and life
sciences, education, communications and telecommunications,
call centers, customer relationship management, public and
government sector administration, defense and intelligence
management, public and private utilities, transportation,
construction and engineering, technology innovation,
insurance, financial transaction processing, analysis and
management, governance, risk and compliance management,
management of supply chains, orders, procurement, inventory,
assets, projects and manufacturing, business process
outsourcing, business consolidation management, business
quality management, business project management, business
stakeholder-shareholder relationship management, and
strategic business, simulation, enterprise and resource
planning; platform as a service (PAAS) featuring software
platforms to collect, manage, analyze, retrieve, monitor,
maintain, report on, structure, model, forecast, present and
display data and information in the fields of marketing,
sales, retail services, customer service, hospitality,
contracts, human resources, clinical research, health care,
health and life sciences, education, communications and
telecommunications, call centers, customer relationship
management, public and government sector administration,
defense and intelligence management, public and private
utilities, transportation, construction and engineering,
technology innovation, insurance, financial transaction
processing, analysis and management, governance, risk and
compliance management, management of supply chains, orders,
procurement, inventory, assets, projects and manufacturing,
business process outsourcing, business consolidation
management, business quality management, business project
management, business stakeholder-shareholder relationship
management, and strategic business, simulation, enterprise
and resource planning; infrastructure as a service (IAAS)
featuring computer software platforms to collect, manage,
analyze, retrieve, monitor, maintain, report on, structure,
model, forecast, present and display data and information in
the fields of marketing, sales, retail services, customer
service, hospitality, contracts, human resources, clinical
research, health care, health and life sciences, education,
communications and telecommunications, call centers,
customer relationship management, public and government
sector administration, defense and intelligence management,
public and private utilities, transportation, construction
and engineering, technology innovation, insurance, financial
transaction processing, analysis and management, governance,
risk and compliance management, management of supply chains,
orders, procurement, inventory, assets, projects and
manufacturing, business process outsourcing, business
consolidation management, business quality management,
business project management, business
stakeholder-shareholder relationship management, and
strategic business, simulation, enterprise and resource
planning.
37.
FILTER CHAIN WITHIN A GATEWAY TENANCY FOR SOFTWARE ASSURANCE IN A CLOUD ENVIRONMENT
Techniques for implementing a filter service for software assurance within a cloud environment are disclosed. At the filter service, a plurality of requests is received from a gateway service operating within a gateway tenancy of a cloud environment. Each request of the plurality of requests is either inbound to a customer tenancy or outbound from the customer tenancy of the cloud environment. Each request is processed by the filter service, by one or more of (i) validating a schema of one or more requests of the plurality of requests, (ii) sampling one or more requests of the plurality of requests, and (iii) auditing one or more requests of the plurality of requests. A first request of the plurality of requests is allowed passage to a corresponding target destination, and a second request of the plurality of requests is denied to a corresponding target destination.
Techniques for implementing a generic gateway service for software assurance within a cloud environment are disclosed. At a gateway tenancy of a cloud environment, a plurality of packets are received from a compute instance operating within a customer tenancy of the cloud environment. A first subset of the plurality of packets are destined for an Internet Protocol (IP) address external to the cloud environment. A second subset of the plurality of packets are destined for a cloud resource within the cloud environment. The first subset of the plurality of packets and the second subset of the plurality of packets are processed at the gateway tenancy and at an application layer (Layer 7). The first subset of the plurality of packets are transmitted from the gateway tenancy to the IP address over a public network. The second subset of the plurality of packets are denied passage to the cloud resource.
Techniques are disclosed for managing the processing of pointers during a concurrent phase of operations using a monotonic state machine. The monotonic state machine defines multiple distinct state designations. During the phase of concurrent operations, the state of any given pointer may be described by a state designation defined by the monotonic state machine. The state designations defined by the state machine have stacking requirements. For example, a third state designation corresponding to a greater level of accessibility necessarily implies a second state designation corresponding to a lesser level of accessibility. The system models changes to a pointer's state during the phase of concurrent operations as monotonic transitions between state designations. If a pointer is transitioned from a former state designation to a latter state designation during the phase of concurrent operations, the pointer is not reverted backwards to the former state designation during the phase of concurrent operations.
Techniques for extracting an entity-specific network from a cross-domain network for query execution are disclosed. The cross-domain network includes multiple terminologies. Connections are generated (a) between terms, represented as nodes, within each terminology and (b) between terms of different terminologies to generate the cross-domain network. The connections are defined by relationships. An entity-specific network is a subset of the cross-domain network. The entity may be a symptom, a disease, or a condition. The system receives a query that includes a term. The system locates the node representing the term in the cross-domain network. The system identifies nodes within “N” hops of the node and generates an entity-specific network comprising the nodes within “N” hops of the node. The system executes the query on the entity-specific network. The query results are presented to the user on an interface, e.g., dashboard.
G16H 70/20 - ICT specially adapted for the handling or processing of medical references relating to practices or guidelines
G16H 10/60 - ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
G16H 40/20 - ICT specially adapted for the management or administration of healthcare resources or facilitiesICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
41.
DYNAMIC TIME SLICE AUTOENCODER NETWORK ANOMALY DETECTION
Techniques are provided for dynamic time slice autoencoder network anomaly detection. In an example method, a computing system receives, by a router, network traffic. The computing system generates, by the router, data characterizing attributes of the network traffic traversing the router. The computing system generates, by the router, representations of the data, wherein each representation of the data is generated for one of a number of predetermined time frames. The computing system inputs each representation of the representations into a unique autoencoder corresponding to the predetermined time frame of the representation, where each unique autoencoder is trained to output a value indicating that the representation contains one of anomalous activity or non-anomalous activity. The computing system identifies a distributed denial of service (DDOS) attack based on the values from the autoencoders. The computing system identifies a source of the DDOS attack.
A policy enforcement system is disclosed that provides a secure framework for accessing resources by evaluating multiple different policy corpuses across different tenancies in a cloud environment. The system obtains a request to perform an action on a resource. The system identifies multiple policies related to a compartment identified for the resource and identifies multiple policies related to a compartment identified for a requestor of the request. The resource resides in a first tenancy and the requestor is associated with a second tenancy in the multi-tenancy cloud computing environment. The system generates an access tree comprising multiple leaf nodes, where the leaf nodes represent the multiple policies. The system evaluates the access tree comprising the multiple policies and provides an evaluation result based on the evaluation.
Unstructured clinical text, such as queries and extracted entities, is semantically normalized to standardized medical codings and schemas using embedding-based similarity search. Upon receiving clinical language input through an application programming interface, a semantic embedding model generates high-dimensional representations of the text. A hybrid search combining vector similarity and keyword matching is performed against a pre-indexed database of medical terminology and semantic object schemas. For broad queries, a ranked list of candidate codings and schemas is provided; for structured inputs containing pre-identified entities, each entity is linked to a ranked set of relevant concepts, with searches filtered by entity type and coding system. These approaches deliver accurate, context-aware mapping of clinical terms to standard vocabularies, enabling downstream clinical analytics, decision support, and data interoperability with high scalability and minimal latency.
G16H 50/70 - ICT specially adapted for medical diagnosis, medical simulation or medical data miningICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
In some embodiments, a system receives, from a requester system, a request for a data integration process that exchanges data between the requester system and a target system and then obtains one or more attributes of the requester system and one or more attributes of the target system. Next, the system may generate a connectivity configuration based on the one or more attributes of the target system using an artificial intelligence (AI) model and generate a transformation configuration based on the one or more attributes of the requester system and the one or more attributes of the target system using the AI model. The system may then execute the data integration process using the connectivity configuration to establish a connection between the requester system and the target system and using the transformation configuration to perform a data transformation process on data transferred between the requester system and the target system.
H04L 41/16 - Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
H04L 41/0806 - Configuration setting for initial configuration or provisioning, e.g. plug-and-play
H04L 67/1095 - Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes
45.
DYNAMICALLY ENABLING/DISABLING CHANGE TRACKING ON RUNNING REPLICATION FOR A MULTI-CLIENT FILE SYSTEM
A change tracking and delta generation (CTDG) system servicing multiple clients is disclosed for enabling the delta generation during replication to be proportional to the number of changes between two snapshots instead of the size of the snapshots. In some embodiments, CTDG can track changes to a file system, where change-tracking can be dynamically switched between two sub-CT modes, file level or directory level. Change-tracking (CT) keys may be generated accordingly and stored in a data structure to be used for delta generation during replications for clients. In some embodiments, change-tracking may be enabled or disabled for different clients. CTDG can coordinate the CT enablement/disablement to ensure replications can use the information in CT keys properly.
A computing device of the control plane may disconnect a server from at least one of a network path or a first boot storage device, the server having an initial network address. The computing device of the control plane may store a server state of the server in a storage device of the control plane. The computing device of the control plane may connect the server to an update storage device containing an update code. The computing device of the control plane may instruct the server to execute the update code. The computing device of the control plane may determine that the server has executed the update code. The computing device of the control plane may restore the server to the server state. The computing device of the control plane may reconnect the server to at least one of the network path or the first boot storage device.
Combining allowlist and blocklist support in data queries includes performing operations including obtaining a runtime query and extracting a set of runtime tuples from the runtime query. The operations further include processing the set of runtime tuples by an allowlist semantic comparator comparing the set of runtime tuples with an allowlist to obtain a first comparison result and by a blocklist semantic comparator comparing the set of runtime tuples with a blocklist to obtain a second comparison result. The blocklist semantic comparator performs an inverse comparison of the allowlist semantic comparator. The operations further include combining the first comparison result with the second comparison to form an access determination and executing the runtime query according to the access determination.
Described herein are various techniques for managing media related to a construction project, which may relate to one or more buildings to be built and/or renovated as part of the construction project. In accordance with techniques described herein, the media may be managed according to information regarding the construction project that is stored by one or more other systems separate from a media management system. The other systems may be construction information management systems that each store information regarding a construction project, including information relating to planning and/or execution of the construction project. For example, the media management system may generate tags based on the information regarding the construction project, may be trained to recognize features in the media that relate to the tags, and add the appropriate tags to the media based on the recognized features.
Techniques for evaluating a language model are disclosed herein. A language model is used to generate responses to queries using documents as context. A model evaluator evaluates a query, document, response, triplet to classify the triplet. The triplets are evaluated using key-phrase extraction, summary generation, cross-encoding and similarity scoring for members of the triplets. The triplets are evaluated by generating a summary of the document and evaluating similarity for the summary and response. The triplets are aggregated and/or clustered to determine one or more classifications or attributes based on the scores. Classifications and/or attributes of triplets are used as training data to train and/or fine tune the language model. A classification of a triplet is used to indicate that a chatbot should switch from a first language model to a second language model for generating a subsequent response to a subsequent query.
Here is artificially intelligent (AI) operations (AIOperon) by a user-facing network element to remedy an operational deficiency of a remote network element in a distributed system such as a data center or a computing cloud. The user-facing network element receives a deficiency text string that indicates an operational deficiency of the remote network element. Based on the deficiency text string and retrieval augmented generation (RAG), a few highly semantically relevant standard operating procedure specifications are contextually selected, and each contains multiple natural language sentences and a few computer commands. A large language model (LLM) processes a linguistic prompt that contains the deficiency text string and the selected standard operating procedure specifications. From the linguistic prompt, the LLM inferentially generates a computer script that contains an inferred sequence of computer commands. Applying the computer script to the remote network element remedies or mitigates the operational deficiency of the remote network element.
H04L 41/5009 - Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
H04L 41/16 - Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
51.
Techniques For Deterministically Terminating A Phase Of Concurrent Operations
Techniques are disclosed for avoiding termination indeterminism associated with a phase of concurrent operations. For instance, during a concurrent marking phase of a garbage collection cycle, the system generates a first data structure to record reachability information of runtime objects in a memory area that are strongly reachable, and the system generates a second data structure to record reachability information of runtime objects that are reachable but not strongly reachable. In particular, the second data structure records reachability information of runtime objects that could become strongly reachable as a result of a potential resurrection. If the potential resurrection occurs during a termination procedure for the concurrent marking phase, the system takes note of the resurrection and proceeds with the termination procedure as normal. The system uses the second data structure to update the reachability information recorded by the first data structure after the termination procedure is completed.
Techniques are described for generating synthetic multilingual text and audio data for training speech models. In one aspect, a computer-implemented method is described that includes: generating, by a first generative artificial intelligence model, keyphrases for a domain, generating entities and associated normalized forms, generating, by a second generative artificial intelligence model, text scripts for the domain based on the keyphrases and the entities and their associated normalized forms, generating, by one or more audio models, audio associated with each of the text scripts, and training, using the text scripts and the audio associated with each of the text scripts, one or more machine learning models.
Example systems, methods, and computer-program products for prioritizing analysis of anomalies in a network for investigation are described. In an example, network data traffic analysis includes accessing a plurality of network traffic measurements along a plurality of dimensions. The network data traffic analysis also includes determining combined and scaled dimension-specific distances for a subset of the network traffic measurements and generating a set of aggregate composite scores for the subset of the network traffic measurements.
Techniques are described for creating a network-link between a first virtual network in a first cloud environment and a second virtual network in a second cloud environment. The first virtual network in the first cloud environment is created to enable a user associated with a customer tenancy in the second cloud environment to access one or more services provided in the first cloud environment. The network-link is created based on network resources and one or more link-enabling virtual networks being deployed in the first cloud environment and the second cloud environment.
Techniques for preventing leakage of sensitive information through a signature are disclosed. Data is received, and the data is transmitted to a signature and encryption (SE) service, along with an encryption key. Signed and encrypted data is received from the SE service, where the signed and encrypted data is (i) encrypted using the encryption key and (ii) signed by the SE service. A verification is performed to verify that that sensitive information (such as the encryption key) is not leaked through a side channel of a signature of the signed and encrypted data, such as by (i) determining a length of the side channel of the signature, and (ii) verifying that the length of the side channel of the signature does not exceed a threshold length. Responsive to verifying that the sensitive information is not leaked through the side channel of the signature, the signed and encrypted data is transmitted.
H04L 9/32 - Arrangements for secret or secure communicationsNetwork security protocols including means for verifying the identity or authority of a user of the system
G06F 21/62 - Protecting access to data via a platform, e.g. using keys or access control rules
G06F 21/64 - Protecting data integrity, e.g. using checksums, certificates or signatures
A computer system includes a detector that monitors DNS communications to detect data exfiltration and/or infiltration that is attempted or has occurred in the DNS communications. DNS communications are stored and parsed to distinguish content that is potentially not public being communicated in a prefix and content contained in a publicly available suffix. The content of the prefix is examined to determine an amount of information conveyed in the prefix based at least in part on a length and a number of unique characters of the prefix. The detector aggregates communications for network sources and/or destinations, and/or for different network groups or characteristics to determine sources, destinations, groups, and/or characteristics associated with aggregate amounts of information that satisfy one or more notification conditions or trigger one or more other corrective actions.
Described are techniques for adverse drug detection using ontology-augmented Large Language Models (LLM). In one aspect, a method is provided that includes accessing medical text associated with a patient. The medical text is annotated for drug and event entities to create labelled medical text, and spans in the labelled medical text are linked to concepts in an ontology. Possible paths in the ontology are identified based on the labeling and linking, each of the possible paths are ranked, and a top number of paths are identified based on the ranking. A verbalized form of each path of the top number of paths are then concatenated with associated portion(s) of the medical text to create ontology augmented text instances and a prompt having the ontology augmented text instance(s) are generated. Thereafter, one or more adverse drug reaction relation predictions are generated, by an LLM, for the patient based on the prompt.
G16H 20/10 - ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to drugs or medications, e.g. for ensuring correct administration to patients
G06F 18/2413 - Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
G16H 10/60 - ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
Systems, computer-readable media, and methods are described for causing display of a map and receiving configuration input to map controls filtering out resources to select a subset of unfiltered resources. The map is displayed in a filtered state showing at least a particular resource of the subset at a particular resource location associated with the particular resource without showing the filtered resources on the map. While the map is in the filtered state, a selection of a particular task item is received and the particular resource is automatically scheduled to complete the particular task item a particular status of the particular task item is automatically updated, a resource task schedule is automatically updated for the particular resource to include the, and the map is automatically updated to include an indication that is based at least in part on both the particular task location and the particular resource location.
A method for intelligent admission control in SBI NFs includes receiving, at an SBI NF operating HTTP server, an SBI request message and determining sender timestamp and a maximum response time for the SBI request message. The method further includes determining, by the SBI NF operating as the HTTP server, a predicted response time for processing the SBI request message, the predicted response time including a predicted time for delivering an SBI response message to the SBI request message to an SBI NF operating as an HTTP client. The method further includes determining whether a sum of a current time and the predicted response time exceeds a sum of the sender timestamp and the maximum response time and aborting processing of the SBI request message when sum of the current time and the predicted response time exceeds the sum of the sender timestamp and maximum response time.
For database high availability without data loss, a storage drive is resilvered in the background while all replicas of a database remain in service. A distributed system persists a first replica of a database in a first storage computer (SC) and a second replica of the database in a second SC. The database contains a database extent that contains a database block. An instance of the database block is persisted in a storage drive of the first SC. When data loss of the database block instance is discovered, the first SC increments a repair count of the database. While the first SC repairs the damaged database extent that includes the lost database block, a second SC detects that the repair count of the database exceeds a repair count in a write request, which causes the write request to be rejected in a way that indicates that the write request should be retried.
G06F 11/20 - Error detection or correction of the data by redundancy in hardware using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
61.
Auxiliary Network Attachments for Containerized Workloads
A system for instantiating distinct network namespaces for each container within a pod is provided. The system includes one or more computer-readable non-transitory storage media embodying software that is operable when executed to instantiate a first network namespace for a pod, determine whether a pod specification includes an indication of an intent to create a second network namespace for the pod, and, responsive to determining that the pod specification includes the indication of the intent to create the second network namespace, instantiate the second network namespace for the pod. The pod includes a logical unit configured to execute one or more containers, in which the pod is managed by a container orchestration system, and the pod specification includes a file including one or more attributes of the pod.
Techniques are described for establishing a private network path from a first cloud environment to a second cloud environment. A tenancy associated with the first cloud environment is provided in the second cloud environment. The tenancy includes a set of one or more resources that enable communication between the first cloud environment and the second cloud environment. A request originating in the second cloud environment and associated with a service provided by the first cloud environment is caused to be received by a first resource from the set of one or more resources. Using at least one resource from the set of one or more resources, the request is transmitted from the second cloud environment to first cloud environment.
A method for intelligent admission control in SBI NFs includes receiving, at an SBI NF operating HTTP server, an SBI request message and determining sender timestamp and a maximum response time for the SBI request message. The method further includes determining, by the SBI NF operating as the HTTP server, a predicted response time for processing the SBI request message, the predicted response time including a predicted time for delivering an SBI response message to the SBI request message to an SBI NF operating as an HTTP client. The method further includes determining whether a sum of a current time and the predicted response time exceeds a sum of the sender timestamp and the maximum response time and aborting processing of the SBI request message when sum of the current time and the predicted response time exceeds the sum of the sender timestamp and maximum response time.
An artificial intelligence model having an audio encoder and a language model decoder is trained through unified cross-modal processing. Audio data and corresponding ground truth outputs are received as inputs. The audio encoder generates encoded audio features from the audio data. An adaptation layer projects the encoded audio features to generate projected features aligned with an embedding space of the language model decoder. The language model decoder processes the projected features to generate model outputs. Training combines an alignment loss between projected features and expected decoder input embeddings with an output loss between model outputs and ground truth outputs. The combined losses form a total loss for updating model parameters. Direct fusion between audio encoding and language model processing is enabled without requiring separate automated speech recognition and large language model components, reducing computational overhead while maintaining accuracy in processing audio inputs.
G10L 25/30 - Speech or voice analysis techniques not restricted to a single one of groups characterised by the analysis technique using neural networks
Techniques for controlling packet flows are described. In an example, a packet is sent on a virtual network. The packet's header includes scoping data that indicates a network boundary within which the packet is permitted and/or prohibited to flow. A network virtualization device of a substrate network receives the packet. The network virtualization device determines the scoping data from the header and, based on network configuration information, determines the forward flow of the packet. If the forward flow falls within a permitted network boundary indicated by the scoping data, the network virtualization device sends the packet forward. Otherwise, the packet is dropped.
In accordance with an embodiment, described herein are systems and methods for use with a computing environment, for providing a determination of model fitness and stability, for model deployment and automated model generation. A model fitness and stability component can provide one or more features that support model selection, use of a model deployability score and deployability flag, and mitigation of model drift risk, to determine model fitness and stability for a particular application. For example, embodiments may be used with analytic applications, data analytics, or other types of computing environments, to provide, for example, a directly actionable risk prediction, in finance applications or other types of applications.
A method for automatically deriving and using a steering of roaming policy includes learning a steering of roaming (SoR) policy of a home public land mobile network (HPLMN). The method further includes receiving, by a proxy function, an SBI message from a first VPLMN and relating to an outbound roaming subscriber. The method further includes determining, by the proxy function and using the SoR policy for the HPLMN and parameters from the SBI message, whether the SoR policy indicates that the HPLMN will steer the outbound roaming subscriber to a VPLMN other than the first VPLMN. The method further includes blocking, by the proxy function, the SBI message when the SoR policy indicates that the HPLMN will steer the outbound roaming subscriber to a VPLMN other than the first VPLMN.
An aspect of the present disclosure facilitates development of integration connectors. In one embodiment, a digital processing system receives a metadata corresponding to a security policy for accessing a target component, the metadata specifying a respective set of actions to be performed for each flow of a pre-defined set of flows for implementing the security policy. The system inspects the metadata to identify the respective set of actions and forms software instructions corresponding to the respective sets of actions. The system then generates an integration connector incorporating the software instructions, such that the integration connector, when operative, implements the security policy in accessing the target component.
Disclosed is an approach to proactively identify software regressions in production cloud environments with saved copies of actual customer workloads. By running real customer workloads, regressions can be detected and the cause of the culprits identified to help facilitate resolution.
A method for automatically deriving and using a steering of roaming policy includes learning a steering of roaming (SoR) policy of a home public land mobile network (HPLMN). The method further includes receiving, by a proxy function, an SBI message from a first VPLMN and relating to an outbound roaming subscriber. The method further includes determining, by the proxy function and using the SoR policy for the HPLMN and parameters from the SBI message, whether the SoR policy indicates that the HPLMN will steer the outbound roaming subscriber to a VPLMN other than the first VPLMN. The method further includes blocking, by the proxy function, the SBI message when the SoR policy indicates that the HPLMN will steer the outbound roaming subscriber to a VPLMN other than the first VPLMN.
H04W 8/02 - Processing of mobility data, e.g. registration information at HLR [Home Location Register] or VLR [Visitor Location Register]Transfer of mobility data, e.g. between HLR, VLR or external networks
Systems, methods, and computer-readable media may provide for contextual alt text generation for web images. A webpage may be parsed to extract images and text. The images may be analyzed to detect representations of persons. Face localization may be performed to detect faces in the representations of the persons. A face embedding for each detected face may be generated to create a set of face embeddings. The text extracted may be analyzed to detect named entities. A name embedding for each detected named entity may be generated to create a set of name embeddings. Bipartite matching may be used to correlate at least part of the set of face embeddings with at least part of the set of name embeddings to create a set of correlation results. Contextualized alt text may be caused to be generated based on the set of correlation results from the bipartite matching.
Embodiments operate a blood bank. Embodiments receive a request for a blood product corresponding to a patient and determine whether the request is a routine request or an emergency request. Embodiments determine an availability of patient history for the patient and display a color coded listing of parameters corresponding to the request, the parameters including a different color for blood product requests when a designated time has elapsed.
G16H 40/20 - ICT specially adapted for the management or administration of healthcare resources or facilitiesICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms
G16H 10/40 - ICT specially adapted for the handling or processing of patient-related medical or healthcare data for data related to laboratory analysis, e.g. patient specimen analysis
G16H 10/60 - ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
G16H 40/67 - ICT specially adapted for the management or administration of healthcare resources or facilitiesICT specially adapted for the management or operation of medical equipment or devices for the operation of medical equipment or devices for remote operation
73.
EVENT-DRIVEN ARCHITECTURE TO ORCHESTRATE ASYNCHRONOUS ACTIVITIES TO KEEP DATA UP-TO-DATE AT AN ADAPTIVELY VARIABLE, TYPE-SPECIFIC CADENCE BASED ON PROCESS CONTROL METADATA
Methods, systems, and computer-readable media are provided for asynchronous process automation that efficiently manages asynchronous processes based on process control metadata to keep data up-to-date at an adaptively variable type-specific cadence based on process control metadata, eliminating the need for manual engagement to maintain operational activities. An event-driven framework is provided for the asynchronous process automation that includes AI agent(s) for interactions, along with a process management system that tracks pending potential work items due to one or more changes in source data. The process management system activates an asynchronous process management process after a wait period. The process management system then determines which pending potential work items to perform in an iteration of evaluation of the pending potential work items. The process control metadata includes stored limitations on a timing for performing individual processes. The process management system prioritizes processes based on process control metadata, such as prioritizing specific ones over others, including throttling certain other processes during high impact time-sensitive activities.
Techniques for enabling a customer operator of a cloud service provider (CSP) the ability to disable operator access to resources in a customer cloud environment are disclosed. Operator access may be disabled or suspended by operators of the CSP customer initiating a disable command. Disabling operator access includes (a) terminating existing sessions that provide operators access to the resources, (b) rejecting new requests for credentials to establish sessions that provide operator access, and/or (c) revoking existing credentials used to establish sessions that provide operator access. Disabling operator access may apply to resources in the customer cloud environment or to a subset of resources and/or may apply to some operators but not to other operators. The operators may be of the same or different categories of operators. At the conclusion of a designated period of time, the ability of operator to access the customer cloud environment may be restored.
A method for implementing a proof-of-work challenge for transmission of data to a non-authenticated gateway is disclosed. The method includes receiving, by the gateway and from a device, a challenge request; and transmitting a proof-of-work challenge to the device. The method further includes receiving, from the device, a solution to the challenge, wherein the solution to the challenge accompanies data. The method further includes verifying a validity of the solution to the challenge; and storing and/or processing the data, responsive at least in part to the solution being valid for the challenge. In an example, the solution to the challenge is to be derived by the device, without an intervention by a user of the device. In an example, the challenge request and the solution to the challenge are received from a library that is packaged with a mobile application being executed within the device.
An object access service is implemented on a computer system for configuring and enforcing multi-dimensional data security using security assignments beyond role-based access controls. The computer system accesses a request submitted on behalf of a user for access to database structure(s). The computer system requests predicate(s) mapped to role(s) assigned to the user and stored in association with the database structure(s). Predicate(s) submitted to the database may be dynamically filled in by the database pursuant to retrieving data from the database structure(s) to reference security assignment field(s) and security assignment value(s) of the security assignment. The predicate(s) are used to retrieve a security assignment that restricts access to the database structure(s) beyond the role(s) assigned to the user. The security assignment is used to restrict data accessible from the database structure(s) to generate a result set, which is transmitted to a client consumer system for consumption via a consumer interface.
Query optimization techniques are provided for graphs with schema that can significantly reduce remote communication (i.e., networking) for distributed graph queries, with a focus on distributed asynchronous traversals. The optimization techniques include (i) a set of rules to infer and add schema information to graph queries, thus enabling better query planning, and (ii) a set of rules for reducing redundant schema information from the query execution plan, thus making the execution more lightweight.
Techniques for responding to multilinguistic queries using an agent core are disclosed herein. An agent core is trained and/or fine-tuned in a first language to generate instructions (i.e., commands) for answering a query. A language identification and/or translation model receives queries, identifies languages associated with the queries, and translates the queries to the first language. The agent core generates instructions in the first language based on the translated queries. The instructions include instructions, in the first language, to perform actions such as retrieval, generation, contextual understanding, or calculation, in both the first language and the second language. The results of executing the instructions, including an action performed in the first language and an action performed in one or more second languages, are combined, ranked and/or reranked to generate an answer to the query.
Various embodiments of the present technology generally relate to systems and methods for providing a modem adjustment engine. In an example, the modem adjustment engine determines a connection event between an endpoint and an access point within a network, where the access point facilitates data transmission and routing between the endpoint and the network. Based on the connection event, the modem adjustment engine determines the current latency of the network for the endpoint and a current MTU of a modem associated with the endpoint. Based on the current MTU, and in some cases the current latency of the network, the modem adjustment engine determines an adjusted MTU. Then, the modem adjustment engine adjusts the current MTU of the modem to the adjusted MTU.
Systems, methods, and other embodiments are described herein which are associated with detection and correction of large language model responses in natural language that contain inaccurate numeric comparisons. In one embodiment, a method includes intercepting an initial response by a large language model to an initial prompt. The method includes detecting that the initial response contains an inaccuracy in a numeric comparison based on a validation check of the numeric comparison. The method includes generating an alternative prompt to the large language model that is configured to cause the large language model to correct the inaccuracy in a rectified response. And, the method includes returning the rectified response by the large language model to the alternative prompt in place of the initial response.
A Video Highlight Summarization System (VHSS) is described for generating a personalized video highlight summary from a video source (e.g., a sport match) based on a user's query. In some embodiments, the VHSS may perform multimodal data analysis. The multimodal data may include information from video, audio, and text from images associated with the video and from user's query. A user may provide a query specifying the user's preferences (e.g., events of interest) and criteria (e.g., summary duration). In some embodiments, encoded embeddings based on the video, audio, text, and the user query may be aligned to enhance similarity search result. A subset (e.g., highlights) of the video clips is selected from the video source by maximizing the summation of scores of highlight clips to best fit the user's preferences while meeting the user's criteria with diverse clips.
G06F 16/783 - Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
82.
Efficient Cryptographic Key Management in Resource Constrained Devices
One or more embodiments address the transition to Post Quantum Cryptography (PQC) within secure element hardware environments. The embodiments focus on key management solutions for resource-constrained devices running JAVA CARD or similar platforms. PQC implementations face challenges from large key sizes that strain secure element resources, including RAM, ROM, flash memory, input/output bandwidth, and processing capabilities. The embodiments present methods for handling PQC keys through importation, exportation, generation, storage, utilization, and protection operations. The implementation manifests as an Application Programming Interface (API) that enables applications to leverage key management capabilities efficiently within secure element resource constraints. The API delivers advantages through memory optimization using flexible condensation and derivation mechanisms. Security benefits emerge from integrating key operations within the certified platform environment, enabling hardware acceleration and side-channel attack countermeasures through native code implementation.
A method implements generating a sampling profile and an instrumentation profile from source code. The method involves identifying a hot subroutine from the source code with the sampling profile. The method further involves executing a hot modifier process with the hot subroutine and one or more of the sampling profile and the instrumentation profile to generate a hot compilation unit. The method further involves executing a cold modifier process with a cold subroutine from the source code and the instrumentation profile to generate a cold compilation unit. The method further involves incorporating the hot compilation unit and the cold compilation unit into a modified compilation schedule.
A Real-Time Enterprise Conversation Analyzer and Decision Enabler (RECA&DE) that enhances the capabilities of enterprise applications, transforming them from transactional or operational support systems into decision support systems. In one aspect, the RECA&DE is capable of performing the following: generating a first graphical user interface for implementing a decision support module within an enterprise application, obtaining data for a discussion event from the enterprise application based on input received via the first graphical user interface, transferring event data to a discussion service, receiving feedback data from an artificial-intelligence platform, the feedback data comprises sentiment data derived by the artificial-intelligence platform based on transcript data generated from conversations of the participants using the discussion service, analyzing the feedback data, and rendering one or more dashboards in a second graphical user interface based on analyzing of the feedback data.
G06Q 10/0637 - Strategic management or analysis, e.g. setting a goal or target of an organisationPlanning actions based on goalsAnalysis or evaluation of effectiveness of goals
Query optimization techniques are provided for graphs with schema that can significantly reduce remote communication (i.e., networking) for distributed graph queries, with a focus on distributed asynchronous traversals. The optimization techniques include (i) a set of rules to infer and add schema information to graph queries, thus enabling better query planning, and (ii) a set of rules for reducing redundant schema information from the query execution plan, thus making the execution more lightweight.
Techniques for mapping proprietary codes with standard codes based on a similarity between network relationships associated with the respective standard codes and proprietary codes are disclosed. The system generates a cross-domain network having sets of terminology, including at least a set of proprietary codes and a set of standard codes. The network includes (a) a plurality of nodes that represent terms in the sets of terminology, (b) inter-terminology connections between nodes across sets of terminology, and (c) intra-terminology connections between nodes within sets of terminology. The system identifies relationships for the terms that define the connections between the nodes. The system generates vector embeddings for the terms by applying a vector embedding function to the relationships and/or the terms associated with each term. Similarity measures are calculated for vector embedding pairs. Pairs having a similarity measure that exceeds a threshold are identified as semantic matches.
G16H 10/60 - ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
87.
SYSTEM AND METHOD FOR A SINGLE LOGICAL IP SUBNET ACROSS MULTIPLE INDEPENDENT LAYER 2 (L2) SUBNETS IN A HIGH PERFORMANCE COMPUTING ENVIRONMENT
Systems and methods for supporting a single logical IP subnet across multiple independent layer 2 subnets in a high performance computing environment. A method can provide, at a computer including one or more microprocessors, a logical device, the logical device being addressed by a layer 3 address, wherein the logical device comprises a plurality of network adapters, each of the network adapters comprising a physical port, and a plurality of switches. The method can arrange the plurality of switches into a plurality of discrete layer 2 subnets. The method can provide a mapping table at the logical device.
Systems, methods, and other embodiments associated with quasi-supervised clustering for activity pattern characterization and anomalous activity detection are described. In one embodiment, a method accesses a plurality of data points. An individual data point of the plurality characterizes a pattern of activity associated with an account. The method splits the plurality of data points into clusters of similar data points. The method evaluates the clusters to detect that the individual data point has changed clusters in a manner indicative of an anomalous change to the pattern of activity. And, the method generates an electronic alert that the pattern of activity has changed anomalously.
H04L 67/1396 - Protocols specially adapted for monitoring users’ activity
G06F 18/23213 - Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
89.
EFFICIENT CRYPTOGRAPHIC KEY MANAGEMENT IN RESOURCE CONSTRAINED DEVICES
One or more embodiments address the transition to Post Quantum Cryptography (PQC) within secure element hardware environments. The embodiments focus on key management solutions for resource-constrained devices running JAVA CARD or similar platforms. PQC implementations face challenges from large key sizes that strain secure element resources, including RAM, ROM, flash memory, input/output bandwidth, and processing capabilities. The embodiments present methods for handling PQC keys through importation, exportation, generation, storage, utilization, and protection operations. The implementation manifests as an Application Programming Interface (API) that enables applications to leverage key management capabilities efficiently within secure element resource constraints. The API delivers advantages through memory optimization using flexible condensation and derivation mechanisms. Security benefits emerge from integrating key operations within the certified platform environment, enabling hardware acceleration and side-channel attack countermeasures through native code implementation.
An aspect of the present disclosure facilitates routing diverse incoming requests to optimal computing options while satisfying requisite performance metrics. In one embodiment, in a computing environment having multiple compute types, a historical data containing characteristics in transport payloads of incoming requests, and characteristics in processing corresponding incoming requests by respective compute types is collected. A system trains, based on the historical data, a machine learning (ML) model to select compute types for incoming requests. Upon receiving a new incoming request sought to be processed, the system extracts from a transport payload of the new incoming request, the first set of characteristics to create a new request context. The system then applies the ML model to the new request context to identify a target compute type and forwards the new incoming request to the target compute type.
The present disclosure describes techniques that facilitate the efficient generation of an application design and an application based on a set of application requirements related to the application. In certain examples, an application generation system is disclosed. The system obtains a set of application requirements for generating an application design for an application. The system generates multiple application designs using a combination of machine language (ML) techniques and Retrieval Augmented Generation (RAG) techniques. The system then evaluates the multiple application designs and selects a particular application design from among the multiple application designs that is deemed to be “best” aligned with the set of application requirements related to the application. As part of generating the multiple application designs and selecting an application design from the multiple application designs, in certain examples, the system provides the selected application design via a user interface of a computing device.
Systems, methods, and other embodiments associated with providing an overview of alignment and risk when applying multiple security frameworks are described. In one embodiment, a method includes accessing (1) a target entity and (2) a control framework having a plurality of controls. The method includes embedding the target entity and the plurality of controls. The method includes quantifying similarities between the embedded target entity and the plurality of embedded controls. The method includes providing the similarities as multivariate input to a regression model that is configured to generate probabilities that individual controls of the plurality are relevant to the target entity. The method includes applying a threshold for relevance to the probabilities to extract a listing of relevant controls that are most relevant to the target entity. And, the method includes generating an electronic alert that includes the listing of relevant controls.
H04L 41/16 - Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using machine learning or artificial intelligence
93.
Concurrency Management Model Using Software Transactional Memory, Dynamic Data Race Detection, And/Or Thread-Local Garbage Collection
Techniques for detecting and preventing potential sources of concurrency issues are disclosed. In response to an attempt to modify a shared data structure, the system generates a copy of the shared data structure in a transaction area allocated for performing the modification through a software memory transaction. The system applies the modification to the copy of the shared data structure, and the system subsequently acquires a lock on the shared data structure. After acquiring the lock, the system checks for accesses to the shared data structure that would conflict with applying the modification to shared data structure. If the check does not reveal a conflicting access, the system commits the software memory transaction by updating the shared data structure to match the state of the copy of the shared data structure. If the check does reveal a conflicting access, the system aborts the software memory transaction.
Techniques for detecting potential concurrency issues and/or causality bugs are disclosed. To this end, the system maintains at least one object graph of event objects. The event objects represent events, and the event objects in an object graph are interconnected by references that represent causal and/or temporal relationships between the corresponding events. Before a thread attempts a task that could feasibly conflict with another task, the system may compel the thread to consult the object graph. For instance, if a task entails accessing a shared object, the system may compel a thread to determine if the last event to mutate the shared object is represented in an object graph of event objects before an event corresponding to the thread's current attempt to complete this task. If the last mutation event is absent from the object graph, the system may prevent or delay the thread from completing the task.
Techniques for managing the distribution of configuration information that supports the flow of packets in a cloud environment are described. In an example, a virtual network interface card (VNIC) hosted on a network virtualization device NVD receives a first packet from a compute instance associated with the VNIC. The VNIC determines that flow information to send the first packet on a virtual network is unavailable from a memory of the NVD. The VNIC sends, via the NVD, the first packet to a network interface service, where the network interface service maintains configuration information to send packets on the substrate network and is configured to send the first packet on the substrate network based on the configuration information. The NVD receives the flow information from the network interface service, where the flow information is a subset of the configuration information. The NVD stores the flow information in the memory.
H04L 41/0813 - Configuration setting characterised by the conditions triggering a change of settings
H04L 41/0853 - Retrieval of network configurationTracking network configuration history by actively collecting configuration information or by backing up configuration information
H04L 45/00 - Routing or path finding of packets in data switching networks
Systems and methods for using InfiniBand routing algorithms for Ethernet fabrics in a high performance computing environment. The method can provide, at a computer comprising one or more microprocessors, a plurality of switches, a plurality of hosts, a topology provider (TP) module, a routing engine (RE) module, and a switch initializer (SI) module. The method can perform a discovery sweep, by the TP, of the plurality of hosts and the plurality of switches and assigns an address to each of the plurality of hosts and the plurality of switches. The method can calculate, by the routing engine, a routing map, based upon a routing scheme, for the plurality of hosts and the plurality of switches, the routing map comprising a plurality of forwarding tables. The method can configure, each of the plurality of switches with a forwarding table of the plurality of forwarding tables calculated by the routing engine.
H04L 41/0813 - Configuration setting characterised by the conditions triggering a change of settings
H04L 41/0895 - Configuration of virtualised networks or elements, e.g. virtualised network function or OpenFlow elements
H04L 41/0897 - Bandwidth or capacity management, i.e. automatically increasing or decreasing capacities by horizontal or vertical scaling of resources, or by migrating entities, e.g. virtual resources or entities
H04L 41/122 - Discovery or management of network topologies of virtualised topologies e.g. software-defined networks [SDN] or network function virtualisation [NFV]
H04L 41/40 - Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks using virtualisation of network functions or resources, e.g. SDN or NFV entities
A computer-implemented method includes receiving an input for a model from a data stream, computing an output from the model, and storing the input and the output as an element of a cache. The method also includes using an algorithm to determine a set of parameters associated with the cache; the algorithm optimizes a function including a time taken by the model to generate outputs from a set of inputs sampled from the data stream. The method further includes calculating a caching score associated with each cache element, based on the set of parameters and the time taken by the model to generate the output, a usage of the element expressed as a number of iterations over which the element has been retained in the cache, and a frequency of usage of the element. The method also includes subsequently removing from the cache the element having the lowest caching score.
A method receives a dataset of records as an input. The method executes a labeling procedure that outputs a matched dataset of the received dataset, the matched dataset including functionally matched pairs of records, where a matching function determines that each functionally matched pair of records represent a respective same entity. The method determines transitive pairs of records including a first record and a second record that belong to a same entity based on the first record and a third record being one of the functionally matched pairs of records representing the same entity and based on the second record and the third record being another of the functionally matched pairs of records representing the same entity. The method labels the transitive pairs of records and functionally matched pairs of records as matched pairs of records based on the records belonging to the same respective entity. The method outputs the labeled matched pairs of records.
A computer-implemented technique for relexicalization of sensitive entities in text data is disclosed. The technique obtains de-identification data identifying sensitive entities from input text and clusters these entities based on their representation of the same real-world things. For each cluster, a representative sensitive entity is determined and used to query a database system. The database returns a best matching candidate sensitive entity based on similarity matching, where each candidate is pre-associated with a relexicalized entity. A large language model (LLM) validates the correspondence between the representative and candidate entities within the input text's context. When validated, the technique generates relexicalized text by substituting cluster entities with the associated relexicalized entity. If validation fails, the technique generates a new relexicalized entity, stores the association in the database, and creates relexicalized text using the generated entity. This approach ensures context-aware, consistent replacement of sensitive entities while maintaining semantic appropriateness through multi-step verification.
Techniques for remediating vulnerabilities in software are disclosed. A system maps third-party software modules to applications in which the third-party software modules are implemented. Upon identifying a vulnerability associated with a particular third-party software module, a system (a) identifies the applications mapped to the third-party software module and (b) initiates a process to remediate the vulnerability in the applications mapped to the third-party software module. The system recommends remediation paths for remediating the vulnerability by applying vulnerability data and mapping data to a generative artificial intelligence (AI) model. If an instance of the third-party software module is stored in the registry, the system scans the instance to determine if the vulnerability exists in the instance.
G06F 21/57 - Certifying or maintaining trusted computer platforms, e.g. secure boots or power-downs, version controls, system software checks, secure updates or assessing vulnerabilities
