A system and method for a hybrid text to speech (TTS) system that receives textual data from a user application; determines that the received textual data is missing from the cache; sends the received textual data to both a remote TTS engine and to a TTS engine in the device; receives speech data from both the remote TTS engine and the TTS engine in the device; and selects or combines, based on a selection policy, the speech data from the remote TTS engine or the TTS engine in the device. The speech data is transmitted to the user application.
G10L 13/08 - Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination
Examples of the present disclosure describe systems and methods determining a root cause of an outage of a dependent service. A method includes detecting an outage of a dependent service, determining a first service dependency of the dependent service, and identifying one or more instances of the first service dependency by accessing a service provider of the first service dependency. The method also includes collecting one or more service level indicators (SLIs) for one or more instances of the first service dependency and determining a health status of the instances of the first service dependency using the SLIs. The method further includes determining a root cause for the outage of the dependent service based on the health status of the instances of the first service dependency.
Embodiments of the present disclosure include techniques for encoding and decoding metadata in error correction codes. During read operation, a decoder generates a first output corresponding to the at least one metadata bit having a first state and a second output corresponding to the at least one metadata bit having a second state. When one of the first and second outputs have a zero value, the decoder sets a value of the at least one metadata bit to the first state or the second state corresponding to the first output or the second output having the zero value. When both the first and second outputs are non-zero, the decoder decodes the codeword with the assumption of both the metadata bit having the first state and the second state to determine if the codeword is correctable with the at least one metadata bit.
H03M 13/11 - Error detection or forward error correction by redundancy in data representation, i.e. code words containing more digits than the source words using block codes, i.e. a predetermined number of check bits joined to a predetermined number of information bits using multiple parity bits
H03M 13/15 - Cyclic codes, i.e. cyclic shifts of codewords produce other codewords, e.g. codes defined by a generator polynomial, Bose-Chaudhuri-Hocquenghem [BCH] codes
4.
NEAR-EYE DISPLAY SYSTEMS UTILIZING AN ARRAY OF PROJECTORS
The present disclosure describes near-eye display systems including an array of projectors and a one-dimensional exit pupil expander. The array of projectors can be arranged along a first dimension and can output image light towards an input coupler within a waveguide that provides one-dimensional exit pupil expansion. In some implementations, arrays of monochromatic projectors are implemented and arranged in offset columns. The input coupler in-couples the image light from the array of projectors into a TIR path within the waveguide. Different optical elements, including diffractive and reflective optics, may be implemented as the input coupler. The image light travels within the waveguide until it interacts with an output coupler. Upon interaction with the output coupler, the image light is expanded in a second dimension transverse to the first dimension and is coupled out of the waveguide.
Layered ingress sharding for multi-tenant services outperforms current sharding techniques, to enhance the reliability and scalability of services. A sharding controller assigns clients and service instances to shards in each of multiple layers. Assignments differ among the layers, at least for clients and may also for service instances. This minimizes adverse effects on clients assigned to a shard with a noisy neighbor, because there are other layers (with a high probability) in which they are not sharing a shard with that noisy neighbor. The sharding controller monitors service instance health and available capacity, which indicates shard health and capacity. Client requests are routed to healthy shards, where retries will eventually find a healthy service instance or, in some examples, requests are routed directly to healthy service instances, eliminating the need for a retry.
H04L 67/1029 - Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers using data related to the state of servers by a load balancer
Disclosed layered sharding for multi-tenant services outperforms current sharding techniques, to enhance the reliability and scalability of services. A sharding controller assigns clients and service instances to shards in each of multiple layers. Each layer can handle requests for any client, with assignments differing among the layers, at least for clients and may also for service instances. This minimize adverse effects on clients assigned to a shard with a noisy neighbor, because there are other layers (with a high probability) in which they are not sharing a shard with that noisy neighbor. As an example, with 40 service instances with 4 per shard, and shuffle sharding assignment with 40C4=91390 shards, the likelihood of a client sharing a shard with the same noisy neighbor in all layers is O(10−8) for two layers, dropping rapidly to O(10−17) for four layers.
The disclosed ingress sharding for multi-tenant services outperforms current sharding techniques, to enhance the reliability and scalability of services. A sharding controller efficiently routes client requests across shards/partitions to improve fault isolation, reduce impact across clients, and distribute loads more evenly. Faults may be isolated within individual shards, and hotspots are reduced to enhance overall system performance. Some examples provide enhanced routing guidance for client requests to reduce reliance on client retry behavior. The sharding controller monitors service instance health and available capacity, which indicates shard health and capacity. Client requests are routed to healthy shards, where retries will eventually find a healthy service instance or, in some examples, requests are routed directly to healthy service instances, eliminating the need for a retry. The underlying sharding arrangement is leveraged to provide well-behaved clients a path to a healthy service instance, whereas the noisy client remains isolated in the affected shard(s).
H04L 67/1029 - Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers using data related to the state of servers by a load balancer
8.
PROVIDING MULTI-REQUEST ARBITRATION GRANT POLICIES FOR TIME-SENSITIVE ARBITRATION DECISIONS IN PROCESSOR-BASED DEVICES
Providing multi-request arbitration grant policies for time-sensitive arbitration decisions in processor-based devices is disclosed. In this regard, a processor-based device provides an arbitration circuit that is configured to select a request tracker entry of a plurality of request tracker entries of a request tracker circuit to apply a multi-request arbitration grant policy. The arbitration circuit determines a count N of a plurality of requests associated with the request tracker entry, and determines a count R of resource elements that are available of a plurality of resource elements of an arbitrated resource. The arbitration circuit determines whether the count R of resource elements that are available is equal to or greater than N, and if so. issues a single arbitration grant for the plurality of requests associated with the request tracker entry to the request tracker circuit.
G06F 13/364 - Handling requests for interconnection or transfer for access to common bus or bus system with centralised access control using independent requests or grants, e.g. using separated request and grant lines
9.
GENERATING UNIVERSAL WEB PROFILES FROM DIVERSE EVENT DATA USING GENERATIVE ARTIFICIAL INTELLIGENCE (AI) MODELS
This disclosure describes a universal user web profile generation system that utilizes one or more generative artificial intelligence (AI) models to generate universal web profiles for users. For example, the universal user web profile generation system uses a combination of neural networks and generative AI models to distill relevant information from the vast amounts and types of user event data and generate relevant universal user event taxonomies. Upon generating the universal user event taxonomies, the universal user web profile generation system can efficiently and accurately generate user profiles based on user web data that aligns with the universal user event taxonomies, ensuring profile compatibility with most or all downstream processes and services that access the user web profiles. Indeed, the universal user web profile generation system generates universal web profiles for users by consolidating extensive user data into a concise and insightful format.
Group-based, intent-aware large language model (LLM) customization is provided. A method includes prompting a first generative model to extract and associate implicit judgments from user responses in real-world conversation logs, the implicit judgments indicating preferred or dis-preferred with a conversation associated with a respective conversation log of the conversation logs, prompting the first or a second generative model to summarize the implicit judgments from the first generative model into generalized preference aspects resulting in group-specific rubrics, the group-specific rubrics indicate significant differences in the generalized preference aspects between groups, and based on the group-specific rubrics from the generative model, (i) augmenting a prompt to a third generative model resulting in an augmented prompt and providing the augmented prompt to the third generative model or (ii) fine-tuning the third generative model, resulting in a group-aligned generative model that provides responses in alignment with a group-specific rubric of the group-specific rubrics.
Examples are disclosed that relate to the use of diffusive autoregression models for generating 3-dimensional (3D) structures of chemical objects. One example provides a method, comprising a) inputting 3D structure data for a chemical object into a trained autoregression transformer model, b) receiving output from the trained autoregression transformer model, the output comprising encoding for a discrete atom type of a predicted next atom to be added to the molecule, c) inputting the 3D structure data for the chemical object and the discrete atom type into a trained diffusion model, d) receiving a position of the predicted next atom from the trained diffusion model, and e) updating the 3D structure data to include the position of the predicted next atom. The method further comprises iterating a), b), c), d), and e) until reaching a stopping criterion and outputting the 3D structure data for the candidate chemical object.
Methods, apparatuses, and products for securing a data lake using artifact-level security, including: storing, in a data lake of a data analytics platform, for one or more data artifacts of a plurality of data artifacts stored in the data lake, artifact-level security data defining permissions to access a corresponding data artifact by one or more roles, wherein the data lake is accessible to a plurality of workloads in the data analytics platform; receiving, by the data lake, a request to access a particular data artifact of the plurality of data artifacts; and controlling access to the particular data artifact using the artifact-level security data for the particular data artifact.
The technology described herein is related to a machine-learning (ML) detection model that detects AI manipulated media, such as videos and images. The technology described herein uses machine learning algorithms to analyze videos and images to generate an authenticity rating. The detection model is trained to understand physics-based constraints. It is difficult to alter an image or generate an artificial image that adheres to the rules of physics in all respects. The detection model can identify possible deviations from the rules of physics in images and videos and use these differences to generate an authenticity metric. Physics-based constraints may include optics-awareness, gravity awareness, material property awareness, conservation of energy, and physical interaction awareness. The detection model is able to identify images and videos that violate the physical constraints. In aspects, the detection model may include a VAE (Variational Autoencoder) and a cGAN (Conditional Generative Adversarial Network) that work together.
G06V 10/44 - Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersectionsConnectivity analysis, e.g. of connected components
G06V 10/82 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
14.
GENERATING AND UTILIZING COMPRESSED GROUNDING DATA FOR SEARCH ENGINES THAT UTILIZE GENERATIVE ARTIFICIAL INTELLIGENCE MODELS
This disclosure describes utilizing a grounding compression system within a search results system to create compressed grounding data to enhance and improve generative search engines (GSEs). For example, the grounding compression system (e.g., a grounding data compression system) dynamically and intelligently reduces large amounts of grounding information into amounts compatible with generative AI models used to answer or provide responses to search queries. Indeed, rather than merely reducing the size of grounding information obtained from a search query, the grounding compression system intelligently distills, condenses, and prunes the grounding data into a compressed block that focuses on the search query, enabling the generative AI model to more efficiently and accurate create a generative response to the search query.
This document relates to using machine learning models to assist users with real world tasks. The disclosed implementations can obtain a demonstration video of a first user performing a real world task. Then, the demonstration video can be processed to obtain augmentation data that can be used at a later time to assist another user with performing the task. For instance, the augmentation data can include keyframes from the demonstration video or captions generated for the keyframes. When another user attempts to perform the task, selected augmentation data can be retrieved and used to prompt a generative model to answer user queries relating to the task.
The technology described herein is related to a machine-learning (ML) detection model that detects AI manipulated media, such as videos and images. The technology described herein uses machine learning algorithms to analyze videos and images to generate an authenticity rating. The detection model is trained to understand physics-based constraints. It is difficult to alter an image or generate an artificial image that adheres to the rules of physics in all respects. The detection model can identify possible deviations from the rules of physics in images and videos and use these differences to generate an authenticity metric. Physics-based constraints may include optics-awareness, gravity awareness, material property awareness, conservation of energy, and physical interaction awareness. The detection model is able to identify images and videos that violate the physical constraints. In aspects, the detection model may include a VAE (Variational Autoencoder) and a cGAN (Conditional Generative Adversarial Network) that work together.
Methods, apparatuses, and products for securing data lake tables using row-level security, including: accessing row-level security data for a table stored in a data lake of a data analytics platform, wherein the row-level security data comprises, for one or more roles, a corresponding expression applicable to the table; identifying, for each role of the one or more roles, a subset of a plurality of rows of the table accessible to a corresponding role by satisfying the corresponding expression; and storing, in the data lake, filtering data identifying, for each role of the one or more roles, the subset of the plurality of rows of the table accessible to the corresponding role.
In network security systems, graph-based techniques may be employed to generate “thumbprints” of security incidents, which may thereafter be used, e.g., for threat actor attribution or the identification of similar incidents. In various embodiments, each security incident is represented by a graph in which security events correspond to nodes, and which encodes associated metadata in additional nodes and/or node/edge attributes. Graph representation learning may be used to compute node and/or edge embeddings, which can then be aggregated into the thumbprint of the incident.
Methods, apparatuses, and products for erasure-coded data transfer using remote direct memory access (RDMA), including: partitioning, by a sender endpoint, a message to be sent to a receiver endpoint into a plurality of data segments by logically subdividing the message into a number of data segments defined as a parameter of an error coding scheme; generating, by the sender endpoint, one or more parity segments from the plurality of data segments by applying the error coding scheme to the plurality of data segments; and sending, by the sender endpoint and to the receiver endpoint, via a plurality of network connections, the plurality of data segments and the one or more parity segments using remote direct memory access (RDMA), wherein each of the data segments and each of the parity segments are sent using RDMA via different network connections of the plurality of network connections.
H04L 67/1097 - Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
H03M 13/37 - Decoding methods or techniques, not specific to the particular type of coding provided for in groups
20.
ADAPTIVE RISK-BASED CHALLENGE SYSTEM USING CONFORMAL UNCERTAINTY CALIBRATION
Aspects of the disclosure include methods and systems for an adaptive risk-based challenge system. A method includes generating challenge features for challenges and generating user features for a user device. The method includes generating, based on the challenge features and the user features, suitability scores for the challenges where the subset of the challenges is selected to meet a suitability threshold, and building, based on the suitability scores, a subset of the challenges. The method includes selecting a challenge from the subset of the challenges and presenting the challenge to the user in order to receive a response, where access to a protected system is determined in accordance with the response.
A system and method for providing coordinated set peripheral device pairing and connection management. Pairing a peripheral device to a host device typically involves presenting an option to a user to consent to connecting a discovered device to the host device. In examples, prior to presenting a connect option, a pairing service obtains product-specific details about the discovered peripheral device and determines whether the device is part of a coordinated set. The details are collected and stored in a local and/or cloud catalog from host devices and/or an original equipment manufacturer of the peripheral device. When the discovered device is determined as a coordinated set, various user interface elements, including a multi-member connect option, are presented to the user. Selection of the multi-member connect option provides consent from the user to connect to all or a subset of the set members via a single user interaction.
Examples of the present disclosure describe systems and methods for on-device, in-browser AI processing. In examples, a selection of an AI pipeline is received. Content associated with the AI pipeline is also received. The content is segmented into multiple data segments and a set of data features is generated for the data segments. AI modules associated with the AI pipeline are loaded to create the AI pipeline. The set of data features is provided to the AI pipeline. The AI pipeline is executed to generate insights for the set of data features. The insights are then provided to a user.
This document relates to using machine learning models to assist users with real world tasks. The disclosed implementations can obtain a demonstration video of a first user performing a real world task. Then, the demonstration video can be processed to obtain augmentation data that can be used at a later time to assist another user with performing the task. For instance, the augmentation data can include keyframes from the demonstration video or captions generated for the keyframes. When another user attempts to perform the task, selected augmentation data can be retrieved and used to prompt a generative model to answer user queries relating to the task.
A server rack door assembly comprises (i) a door panel comprising a ventilated region configured to permit airflow through the door panel; (ii) one or more pressure sensors configured to collect pressure data indicating pressure drop through the door panel; and (iii) a louver system connected to the door panel and comprising: (1) a plurality of slats; (2) one or more actuators in mechanical communication with the plurality of slats and configured to adjust one or more tilt angles for the plurality of slats; and (3) one or more controllers configured to: generate or receive one or more target tilt angles for the plurality of slats, determined based at least on the pressure data indicating pressure drop through the door panel; and control the one or more actuators to adjust the one or more tilt angles for the plurality of slats to the one or more target tilt angles.
A tray is adapted for insertion within a tray slot formed in a chassis. The tray includes a lever and a moveable connector that protrudes from and engages with a side plane of the chassis in response to rotation of the lever.
The present disclosure relates to systems, methods, and computer-readable media for utilizing a new memory allocation function library called PmemMalloc. For example, the PmemMalloc library allocates pre-allocated, partitioned, and fixed shared memory blocks. In addition, by utilizing the PmemMalloc library, the memory allocation system described herein overcomes problems with persistence and enumeration that encumber existing malloc libraries. Indeed, the PmemMalloc library enables the memory allocation system to perform servicing computation in parallel across multiple CPU cores/threads, distribute computation equally among threads, prioritize servicing, among other improvements. Notably, the PmemMalloc library provides major constructs (e.g., persistence, enumeration, and debuggability) not available existing malloc libraries. Additionally, as detailed in this disclosure, the PmemMalloc library migrates various computations out of application-based packet processing to memory block-based deferred enumeration, which improves both packet processing and efficient use of CPU cores on a computing device.
A hardware interface includes a printed circuit board (PCB), including a plurality of PCB dog-bone structures electrically connecting a plurality of PCB vias to a corresponding plurality of socket pins. Each PCB dog-bone structure includes a respective PCB, an intermediary trace, and a socket interface pad electrically connected to a respective socket pin. The plurality of PCB dog-bone structures includes first and second PCB dog-bone structures, which respectively include first and second PCB vias and socket interface pads. The first socket interface pad is adjacent to the second socket interface pad, and the first PCB via is adjacent to the second PCB via. The first socket interface pad and the second socket interface pad are arranged parallel to a first direction, and the first PCB via and the second PCB via are arranged parallel to a second direction, different from the first direction.
A field programmable gate array including a configurable interconnect fabric connecting logic blocks implementing a circuit to: receive input data including data values organized into rows and columns, each row having N data values; select R[i] unmasked data values of a row of the input data in accordance with a mask and an index i of the row; select N−R[i] unmasked data values of another row of the input data in accordance with the mask and an index of the another row; merge the R[i] unmasked data values of the row and the N−R[i] data values of the another row into a combined data vector of N data values; and compute R[i] normalized values based on the R[i] unmasked data values of the combined data vector and N−R[i] normalized values based on the N−R[i] data values of the combined data vector to generate N normalized data values.
G06F 30/331 - Design verification, e.g. functional simulation or model checking using simulation with hardware acceleration, e.g. by using field programmable gate array [FPGA] or emulation
G06F 9/30 - Arrangements for executing machine instructions, e.g. instruction decode
G06F 18/214 - Generating training patternsBootstrap methods, e.g. bagging or boosting
29.
System and Method for Generating a Custom Operating System and Software Development Kit Package
A method, computer program product, and computing system for developing a custom operating system (OS) and software development kit (SDK) package. A feature of an application is determined and a subset of operating system (OS) software components of a plurality of OS software components available for the OS is identified, the subset of OS software components being required to implement the feature. An OS is assembled with the subset of OS software components. A subset of software development kit (SDK) software components from a plurality of SDK software components available for a SDK, the subset of SDK software components being associated with the subset of OS software components. The SDK is assembled with the subset of SDK software components and a software package including the OS and the SDK is generated.
Systems and methods are provided for implementing data backup and recovery using cache-coherent interconnect node-based non-volatile memory. A cache-coherent interconnect node partitions a memory pool into a plurality of memory regions as well as a backup storage into a plurality of memory portions, and pre-allocates a memory region and a corresponding memory portion to each compute node. When a rack-level power loss occurs, and a battery-based power source is activated, a cache-coherent interconnect controller saves data from each memory region into the corresponding memory portion, and subsequently saves an entry for each memory portion in an index portion of the backup storage. Subsequently, the controller causes a power circuitry to shut down the backup power source. After rack-level power restoration and memory region initialization, the controller restores, for each memory region, the data saved in a corresponding memory portion into that memory region, based on information in a corresponding entry.
This document relates to processing of source code using generative language models. One example method includes accessing source code, processing the source code to identify entities in the source code, and generating a graph having nodes representing the entities in the source code and edges representing relationships among the entities. The example method also includes prompting a generative language model to generate augmentation data for the entities based at least on the relationships, receiving, from the generative language model, generated augmentation data, and generating an augmented graph by associating the generated augmentation data with respective nodes of the graph. The augmented graph provides a basis for subsequent operations on the source code by the generative language model.
Methods, apparatuses, and products for erasure-coded data transfer using remote direct memory access (RDMA), including: partitioning, by a sender endpoint, a message to be sent to a receiver endpoint into a plurality of data segments by logically subdividing the message into a number of data segments defined as a parameter of an error coding scheme; generating, by the sender endpoint, one or more parity segments from the plurality of data segments by applying the error coding scheme to the plurality of data segments; and sending, by the sender endpoint and to the receiver endpoint, via a plurality of network connections, the plurality of data segments and the one or more parity segments using remote direct memory access (RDMA), wherein each of the data segments and each of the parity segments are sent using RDMA via different network connections of the plurality of network connections.
H04L 1/00 - Arrangements for detecting or preventing errors in the information received
G06F 13/28 - Handling requests for interconnection or transfer for access to input/output bus using burst mode transfer, e.g. direct memory access, cycle steal
H03M 13/37 - Decoding methods or techniques, not specific to the particular type of coding provided for in groups
A heat-sensing touch interface that profiles the temperature of touch data to identify user input is disclosed herein. Heat profiling touch input improves touch accuracy and allows users to interact with touch devices more naturally. Palm touches can be classified as non-user input, even when appearing as fragmented touches resembling finger touches, when thermal data associated with the palm touchpoint(s) exceeds the temperature range profile for finger touches. Moisture can be classified as non-user input when thermal data associated with the moisture touchpoint(s) is below the temperature range profile for finger touches. The temperature range of user input can be dynamically adjusted. Energy is conserved by activating or sampling a heat sensor array based on detection of touch data. Energy is also conserved by more accurately classifying touch inputs, resulting in reporting and processing fewer non-user inputs. Feedback can be provided at touchpoints classified as user input.
A data management coprocessor, and a method in a data management coprocessor, for interoperating with an artificial intelligence (Al) accelerator and a central processing unit (CPU) in a computer system. The method includes allocating a cache buffer in a memory distinct from the coprocessor. The method also includes predicting a subset of large language model (LLM) weights necessary for generating a subsequent token by an LLM executing in the Al accelerator. The method also includes initiating the caching of these predicted LLM weights into the cache buffer, e.g., before the LLM generates the next token. The data management processor may also roll back a state of the LLM executing in the Al accelerator when a confidence score for the next token meets a criterion indicating a misprediction.
A method implemented in an artificial intelligence (AI) offload die within a system-in-a-package involves hybrid inferencing of an AI model by a remote computing system and a compute die in the system-in-a-package. The method includes identifying a portion of the AI model for use by the compute die, utilizing a network controller in the AI offload die to fetch this portion from the remote computing system, and communicating it to the compute die. Additionally, the network controller in the AI offload die synchronizes AI model inferencing state between the compute die and the remote computing system, ensuring coordinated hybrid AI model inferencing. This approach facilitates efficient distribution and execution of AI tasks between the compute die and the remote computing system, enhancing computational performance and resource utilization.
A device is equipped with a public/private key pair. The private key is stored in a secure location on the device and the public key is utilized to track ownership of the device by a manufacturer, vendor, and/or one or more provisioning services. When a user purchases the device, a transaction involving the public key associated with the device and the user is recorded. The one or more provisioning services, which are provided access to user information, prepare a configuration payload for the device specific to the user and the device. The configuration payload is encrypted using the device's public key. When the device is powered on, the configuration payload is sent to the device. The device decrypts the configuration payload using the device's private key and adjusts one or more configuration parameters based on the configuration payload.
G06Q 20/40 - Authorisation, e.g. identification of payer or payee, verification of customer or shop credentialsReview and approval of payers, e.g. check of credit lines or negative lists
This document relates to processing of source code using generative language models. One example method includes accessing source code, processing the source code to identify entities in the source code, and generating a graph having nodes representing the entities in the source code and edges representing relationships among the entities. The example method also includes prompting a generative language model to generate augmentation data for the entities based at least on the relationships, receiving, from the generative language model, generated augmentation data, and generating an augmented graph by associating the generated augmentation data with respective nodes of the graph. The augmented graph provides a basis for subsequent operations on the source code by the generative language model.
The description relates to thermal management and ensuring adequate cooling of individual computing devices where multiple computing devices operate in proximity to one another. One example can obtain sensed conditions within a rack containing multiple computing devices. The rack receives cooling air from a centralized cooling system. The example can control delivery of make-up cooling air to individual computing devices within the rack to augment the cooling air from the centralized cooling system to remove thermal loads from the individual computing devices.
A computing system receives computer-readable hardware code. A pull request is received that indicates the purpose of the hardware code or updates made to the hardware code. An artificial intelligence (AI) planner agent is invoked that is configured to interact with a plurality of AI review agents that are each trained to: parse the hardware code; analyze the hardware code with respect to the pull request and an attribute; and generate updates to the hardware code in accordance with criteria for meeting the attribute.
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
G06F 8/71 - Version control Configuration management
This disclosure describes utilizing a grounding compression system within a search results system to create compressed grounding data to enhance and improve generative search engines (GSEs). For example, the grounding compression system (e.g., a grounding data compression system) dynamically and intelligently reduces large amounts of grounding information into amounts compatible with generative AI models used to answer or provide responses to search queries. Indeed, rather than merely reducing the size of grounding information obtained from a search query, the grounding compression system intelligently distills, condenses, and prunes the grounding data into a compressed block that focuses on the search query, enabling the generative AI model to more efficiently and accurate create a generative response to the search query.
The present disclosure proposes a method, apparatus and computer-readable medium for sequential recommendation based on cross-domain behavior data. A target user representation of a target user may be generated based on a historical content item sequence of the target user. A cross-domain behavior sequence set may be extracted from a log of a network application. A cross-domain sequence representation set corresponding to the cross-domain behavior sequence set may be generated. A similar sequence representation set similar to the target user representation may be retrieved from the cross-domain sequence representation set. An interaction probability set of the target user interacting with a candidate content item set may be predicted based on the target user representation and the similar sequence representation set.
Systems and methods are provided for accessing a machine learning model configured as a zero-shot cross-lingual text-to-speech model which has been previously trained on a text-to-speech training dataset comprising different bilingual speech transcription pairs, obtaining a first text prompt in a first language, a second text prompt in a second language, a speech sample comprising audio data from an unseen target speaker, providing the first text prompt in the first language, the second text prompt in the second language, and the speech sample from the target speaker as inputs to the machine learning model, and finally, generating a personalized speech output based on the inputs and by at least converting the second text prompt in the second language using a synthesized voice of the target speaker based on the speech sample from the target speaker.
Some embodiments enhance the security of domain name resolution and other DNS operations, by automatically intercepting the DNS operation, determining an associated device identity or ascertaining an associated user identity, and enforcing a security policy based on at least the DNS operation and based on at least one of the identities. Some securable DNS operations include resolution requests, reverse lookups from IP addresses to domain names, DNS record accesses, mail server mappings, redirection, forwarding, and DNS record cache operations. Enforcing the policy includes, e.g., preventing a result requested by the DNS operation, permitting computational progress toward the requested result, allowing a different result, modifying a DNS record, or flushing a DNS record from a cache. In some embodiments, DNS operation security functionality utilizes or implements a conditional access security functionality, thereby providing, e.g., a secure conditional domain name resolution.
H04L 61/4511 - Network directoriesName-to-address mapping using standardised directoriesNetwork directoriesName-to-address mapping using standardised directory access protocols using domain name system [DNS]
This document relates to automated generation of machine learning models, such as neural networks. One example system includes a hardware processing unit and a storage resource. The storage resource can store computer-readable instructions cause the hardware processing unit to perform an iterative model-growing process that involves modifying parent models to obtain child models. The iterative model-growing process can also include selecting candidate layers to include in the child models based at least on weights learned in an initialization process of the candidate layers. The system can also output a final model selected from the child models.
The description relates to resource aware object detection for encoded video streams that can identify frames of the video stream that include an object of interest, such as a human, without decoding the frames.
H04N 19/177 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being a group of pictures [GOP]
H04N 19/169 - Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
46.
USER INTERACTION AND TASK MANAGEMENT USING MULTIPLE DEVICES
The present disclosure provides systems and methods for user interaction and task completion using multiple devices. A set of devices may be used to perform a task, such that different devices may perform different steps of the task. A device management service may update state information at each device of the set, thereby enabling a user to interact with any of the computing devices to perform the task. A device management service may also automatically determine which device should be used by the user, based on task or step requirements, device characteristics, and device capabilities, among other examples. Thus, rather than being required to continue a task on the same device (even when the device is not well-suited for the current step or task), the user is provided with the option to use and, in some instances, is automatically transitioned to use, different devices within the set.
H04L 67/60 - Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
H04L 67/10 - Protocols in which an application is distributed across nodes in the network
47.
MULTIVARIATE THREAT DETECTION FOR A CI/CD PIPELINE
Example solutions protect a continuous integration/continuous deployment (CI/CD) pipeline. Examples collect data from a CI/CD pipeline execution data source and/or a CI/CD pipeline task data source. Based on the collected data, a feature group comprising a plurality of records is created. Each record in the feature group represents an execution of the CI/CD pipeline. An anomaly score is generated, using a model representing historical feature groups, for the feature group representing the execution of the CI/CD pipeline. If the anomaly score is above a threshold, an alert is generated to indicate that the collected data represents an anomalous activity.
G06F 21/55 - Detecting local intrusion or implementing counter-measures
G06F 21/54 - Monitoring users, programs or devices to maintain the integrity of platforms, e.g. of processors, firmware or operating systems during program execution, e.g. stack integrity, buffer overflow or preventing unwanted data erasure by adding security routines or objects to programs
48.
ADAPTIVE AUDIO BASED ON USER PREFERENCES THROUGH LEVERAGING GENERATIVE ARTIFICIAL INTELLIGENCE
Technology is disclosed for adapting a dynamic audio stream to modify sounds of a sound class based on user preferences in systems including gaming systems. A user may provide, via a graphical user interface, a selection of a sound class and a transformation type. Using a pretrained generative artificial intelligence (AI) model, the system may process the dynamic audio stream (e.g., for a gaming instance) to transform instances of the sound class with the transformation type. The output stream from the generative AI model can be used as the audio output. The described technology allows for processing and transforming all dynamic audio streams on a system without specific programming for a particular application or game.
A63F 13/54 - Controlling the output signals based on the game progress involving acoustic signals, e.g. for simulating revolutions per minute [RPM] dependent engine sounds in a driving game or reverberation against a virtual wall
A63F 13/67 - Generating or modifying game content before or while executing the game program, e.g. authoring tools specially adapted for game development or game-integrated level editor adaptively or by learning from player actions, e.g. skill level adjustment or by storing successful combat sequences for re-use
A63F 13/79 - Game security or game management aspects involving player-related data, e.g. identities, accounts, preferences or play histories
49.
DATA MANAGEMENT COPROCESSOR FOR THE SPECULATIVE INFERENCE OF A LARGE LANGUAGE MODEL
A data management coprocessor, and a method in a data management coprocessor, for interoperating with an artificial intelligence (AI) accelerator and a central processing unit (CPU) in a computer system. The method includes allocating a cache buffer in a memory distinct from the coprocessor. The method also includes predicting a subset of large language model (LLM) weights necessary for generating a subsequent token by an LLM executing in the AI accelerator. The method also includes initiating the caching of these predicted LLM weights into the cache buffer, e.g., before the LLM generates the next token. The data management processor may also roll back a state of the LLM executing in the AI accelerator when a confidence score for the next token meets a criterion indicating a misprediction.
G06F 12/0875 - Addressing of a memory level in which the access to the desired data or data block requires associative addressing means, e.g. caches with dedicated cache, e.g. instruction or stack
Systems and methods herein provide a phishing detection engine and its related functions. In an aspect, a phishing detection engine captures focal content displayed via a user interface on a client device. From the focal content, the phishing detection engine extracts features. These features include textual elements and visual elements. Using the features, and in some cases historical user interactions associated with the client device, the phishing detection engine determines whether the features indicate potential phishing activity. If potential phishing activity is detected from the features, the phishing detection engine performs one or more security actions to limit damage of the potential phishing activity, such as blocking execution of an activation step of the phishing activity. In scenarios where the phishing activity is indeterminate, the phishing detection engine may continue to monitor the user's content interaction and extract features from subsequent contents, until a determinate conclusion is reached.
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
This document relates to processing of source code using generative language models. One example method includes accessing source code, processing the source code to identify entities in the source code, and generating a graph having nodes representing the entities in the source code and edges representing relationships among the entities. The example method also includes prompting a generative language model to generate augmentation data for the entities based at least on the relationships, receiving, from the generative language model, generated augmentation data, and generating an augmented graph by associating the generated augmentation data with respective nodes of the graph. The augmented graph provides a basis for subsequent operations on the source code by the generative language model.
09 - Scientific and electric apparatus and instruments
35 - Advertising and business services
42 - Scientific, technological and industrial services, research and design
Goods & Services
Downloadable software in the nature of a mobile application;
downloadable computer software that enables users to access
and interact with information and databases; downloadable
computer software for collecting, editing, organizing,
modifying, bookmarking, storing, sharing and publishing data
and information; downloadable computer software for
uploading, managing, tracking, and sharing customized
content; downloadable computer software for searching,
accessing, displaying, sharing and reviewing newsletters,
research reports, blogs, and articles; downloadable computer
software featuring multimedia content; downloadable computer
software for enabling transmission of images and audiovisual
and video content; downloadable computer software for use in
creating, downloading, uploading, designing, modifying,
reproducing, transmitting, and sharing images, graphics,
fonts, photographs, text, videos, and data; downloadable
recreational game software; downloadable mobile applications
for interactive and recreational games; downloadable
software for games and social networking; downloadable
logic, word, trivia, and puzzle game software via a global
computer network and wireless devices; downloadable
electronic publications in the nature of newsletters,
research reports, articles and white papers on topics of
professional interest; downloadable computer software
development tools; downloadable computer software that
provides web-based access to applications and services
through a web-operating system or portal interface;
downloadable computer software for use in business analytics
and database management; downloadable computer software for
social media, marketing, merchandising, customer service,
website performance, search engine optimization, technology,
consumer goods, retail, and manufacturing; downloadable
computer software for tracking and analyzing user
interaction with customized content; downloadable education
software; downloadable computer software for providing
online courses, seminars, interactive classes, educational
instruction, and course materials; downloadable computer
software for providing access to Internet search engines
featuring information for obtaining job listings, resume
postings, and other job searches; downloadable job
searching, sourcing and recruiting software using artificial
intelligence (AI) for users on a social networking,
employment, and business networking communication platform;
downloadable chatbot software using artificial intelligence
(AI) for users on a social networking, employment, and
business networking communication platform; downloadable
writing and communication software using artificial
intelligence (AI) for assisting platform users with
employment, job sourcing and recruiting, lead generation,
and business-related inquiries; content creation software
using artificial intelligence for users on a social
networking, employment, and business networking
communication platform; downloadable computer software using
artificial intelligence (AI) for employee training and
professional development; downloadable computer software
using artificial intelligence (AI) for providing online
courses, seminars, interactive classes, educational
instruction, and course materials; downloadable podcasts in
the field of in the field of employment, recruitment of
personnel, careers, job resources and listings, and
professional networking and wide field of topics. Providing online employment information and employment
services; providing online business networking services;
providing online career networking services; recruitment and
placement services; providing online employment counseling,
career placement services, and personnel recruitment;
providing an online searchable databases and interactive
databases featuring employment and career opportunities
(term considered too vague by the International Bureau
pursuant to Rule 13 (2) (b) of the Regulations); providing
online information in the fields of employment, recruitment
of personnel, careers, job resources and listings, career
development, professional networking, and employment
advertising; providing recruitment and employment
information, employment advertising, job listings, career
information and advice via an online interactive computer
database from a global computer network; providing an online
artificial intelligence (AI) enhanced searchable database
featuring employment and career opportunities and business,
employment and professional queries and answers (term
considered too vague by the International Bureau pursuant to
Rule 13 (2) (b) of the Regulations); business research and
survey services utilizing artificial intelligence; providing
artificial intelligence (AI) enhanced online computer
databases and online searchable databases in the fields of
marketing, lead generation, sourcing, recruiting, and
business and professional networking (term considered too
vague by the International Bureau pursuant to Rule 13 (2)
(b) of the Regulations); online business networking services
featuring artificial intelligence (AI) solutions;
advertising services; marketing services; marketing
consulting services; advertising, marketing, and promotion
services for businesses; providing advertising and
advertisement services; providing marketing and advertising
solutions for marketing campaigns across a wide range of
industries; providing resources in the nature of online
resource guides for creating advertising and marketing
campaigns that meet business specific business and B2B
needs; creating, placing, displaying, targeting and
disseminating online advertisements for others; providing a
web site which features advertisements for the goods and
services of others on a global computer network (term
considered too vague by the International Bureau pursuant to
Rule 13 (2) (b) of the Regulations); providing advertising
and marketing services via an online platform featuring
sponsored ad content, sponsored ad messaging, text ads,
dynamic ads, and ad placements; providing online advertising
on a computer network; providing business and business
networking information; advertising and marketing services
rendered using artificial intelligence (AI); lead generation
activities and services; advertising and marketing services
in the nature of accessing, extracting, and organizing
information from the Internet and other sources regarding
people, companies, products, marketing, industries and other
categories; lead generation services rendered using
artificial intelligence; employment recruiting services;
professional, staff, personnel and talent recruiting
services; providing an online searchable database featuring
employment and career opportunities and business information
(term considered too vague by the International Bureau
pursuant to Rule 13 (2) (b) of the Regulations); providing
an online searchable database featuring business, employment
and professional queries and answers (term considered too
vague by the International Bureau pursuant to Rule 13 (2)
(b) of the Regulations); providing information online
regarding recruiting and talent solutions (term considered
too vague by the International Bureau pursuant to Rule 13
(2) (b) of the Regulations); providing an online searchable
database featuring professional queries and answers
concerning staffing and hiring information (term considered
too vague by the International Bureau pursuant to Rule 13
(2) (b) of the Regulations); charitable services, namely,
promoting public awareness about charitable, philanthropic,
community service, humanitarian activities and volunteer
activities; providing online career networking services and
information in the fields of employment, recruitment, job
resources, job listings and career path suggestions;
providing business information; providing a web site
featuring business information in the form of audio, video,
transcripts, and other educational materials (term
considered too vague by the International Bureau pursuant to
Rule 13 (2) (b) of the Regulations); providing information,
news and commentary in the field of business; promotion
services for businesses. Providing temporary use of on-line non-downloadable
software; providing temporary use of non-downloadable
software for business and social networking, employment,
careers and recruiting via a website; application service
provider (ASP) services; providing an online software
platform; providing temporary use of on-line
non-downloadable software that enables users to access and
interact with information and databases; providing
customized web pages featuring user-defined information,
audio, text, video, and images (term considered too vague by
the International Bureau pursuant to Rule 13 (2) (b) of the
Regulations); hosting an interactive website featuring
technology that allows users to create, download, upload,
design, modify, reproduce, transmit, and share images,
graphics, fonts, photographs, text, videos, and data;
providing temporary use of on-line non-downloadable software
for collecting, editing, organizing, modifying, bookmarking,
storing, sharing and publishing data and information;
providing temporary use of on-line non-downloadable software
for uploading, managing, tracking, and sharing customized
content; providing temporary use of on-line non-downloadable
software for searching, accessing, displaying, sharing and
reviewing newsletters, research reports, blogs, and
articles; hosting a website for providing general and
customized information in a wide variety of fields, namely,
business, social networking, employment, careers and
recruiting; hosting a website for providing general and
customized information relating to business, current events,
education, entertainment, technology, culture,
entrepreneurship, leadership, management, marketing,
recruiting, career, and professional development; providing
temporary use of on-line non-downloadable software featuring
multimedia content; providing temporary use of on-line
non-downloadable software for enabling transmission of
images and audiovisual and video content; providing
temporary use of on-line non-downloadable software for use
in creating, downloading, uploading, designing, modifying,
reproducing, transmitting, and sharing images, graphics,
fonts, photographs, text, videos, and data; providing
temporary use of online non-downloadable recreational game
software; providing temporary use of online non-downloadable
software for interactive games and recreational game playing
purposes; providing temporary use of online non-downloadable
software for games and social networking; providing
temporary use of online non-downloadable logic, word,
trivia, and puzzle game software; providing temporary use of
non-downloadable computer software featuring electronic
publications in the nature of newsletters, research reports,
articles and white papers on topics of professional interest
in the field of business, social networking, employment,
careers and recruiting via a website; providing temporary
use of on-line non-downloadable software development tools;
providing temporary use of on-line non-downloadable software
that provides web-based access to applications and services
through a web-operating system or portal interface;
providing temporary use of on-line non-downloadable software
for use in business analytics and database management;
providing temporary use of on-line non-downloadable software
for social media, marketing, merchandising, customer
service, website performance, search engine optimization,
technology, consumer goods, retail, and manufacturing;
providing temporary use of on-line non-downloadable software
for tracking and analyzing user interaction with customized
content; providing an online education software platform;
providing temporary use of on-line non-downloadable software
for providing online courses, seminars, interactive classes,
educational instruction, and course materials; providing an
online software platform for employee training and
professional development; providing temporary use of on-line
non-downloadable software for providing access to Internet
search engines featuring information for obtaining job
listings, resume postings, and other job searches; providing
an online software platform for employee training and
professional development that allows users to upload,
manage, and share customized content, access online courses
and content, receive data analytics and insights on learning
and skills development, host online web facilities, links,
webcasts and podcasts for managing and sharing online
content; providing non-downloadable job searching, sourcing
and recruiting software using artificial intelligence (AI)
for users on a social networking, employment, and business
networking communication platform; providing
non-downloadable chatbot using artificial intelligence (AI)
for users on a social networking, employment, and business
networking communication platform; providing
non-downloadable writing and communication online software
using artificial intelligence (AI) for assisting platform
users with writing, communicating, and with employment, job,
recruiting, lead generation, and business-related inquiries;
providing non-downloadable content creation online software
using artificial intelligence for users on a social
networking, employment, and business networking
communication platform; providing non-downloadable software
using artificial intelligence (AI) for employee training and
professional development; providing non-downloadable online
computer software using artificial intelligence (AI) for
providing online courses, seminars, interactive classes,
educational instruction, and course materials; providing
non-downloadable software platform tools for creating,
placing, displaying, controlling and tracking advertising
and marketing content; providing non-downloadable software
platform tools for use in customer relationship management
(CRM), lead generation activities and services, and
tracking, accessing, extracting and organizing sales
information; providing temporary use of a non-downloadable
computer software for providing certification of job skill
assessments online; provision of online non-downloadable
software tools for testing, analysis and evaluation of the
knowledge, skills and abilities of others for job and
employment skills in the field of business, social
networking, employment, careers and recruiting utilizing
artificial intelligence (AI); hosting digital content on
Internet.
The description relates to providing meaningful information relating to a dataset, especially a dynamic dataset that changes over time. One example can obtain text chunks of the dataset grouped by period and extract concepts from the text chunks by period. The example can induce the extracted concepts into a graph structure and detect period communities in the graph structure of individual periods. The example can create period summaries from the detected period communities and determine whether a user query relates to specific periods and/or communities. Where the user query relates to specific periods and/or communities, the example can obtain text answers by mapping the query over relevant period text chunks or relevant period community summaries. The example can obtain a final answer for the user query from the obtained text answers.
Examples are disclosed relating to a circuit for controlling voltage overshoot in a computing system. In one example, a circuit comprises a network of shunt devices arranged into a plurality of branches. Each branch of the plurality of branches includes shunt device(s) connected to an enable pin associated with the branch. Each shunt is configured to induce current through a transistor connected between a power node and a ground node when the shunt device is activated. The circuit comprises a controller connected to a plurality of enable pins corresponding to the plurality of branches of the network. The controller is configured to receive a computing processor voltage, generate a difference value indicating a difference between the processor voltage and a reference voltage, and send enable signal(s) to enable pin(s) to activate the shunt devices based at least on the difference value.
Methods, apparatuses, and products for accelerating container initiation in production environments, including: identifying, based on one or more input/output (I/O) operations associated with a container that are issued in the production environment, a one or more data extents that is sufficient for a host operating system to initiate the container; and responsive to a request to initiate the container, providing the one or more data extents, wherein the host operating system can initiate the container based on the provided one or more data extents without additional portions of a complete dataset for the container being provided to the host operating system.
Methods, apparatuses, and products for packet loss detection in multipath networks, including: encoding, by a source endpoint of a multipath network connection, into each packet of a plurality of packets, an entropy value and a path-specific sequence number, wherein the entropy value is included in a plurality of entropy values each corresponding to a different path of a plurality of paths of the multipath network connection, and wherein the path-specific sequence number comprises a next value in a sequence of values for each subset of the plurality of packets sharing a same network path; sending, by the source endpoint and to a destination endpoint of the multipath network connection, the plurality of packets via the plurality of paths; and performing, by the destination endpoint, packet loss detection based on the entropy value and the path-specific sequence number for the plurality of packets.
This disclosure describes a framework for analyzing dubbed audio segments (audio translations converted into translated speech) of videos where the dubbed audio segments are generated in real time, including being generated locally on a client device. For instance, this disclosure describes a video dubbing system that utilizes various lightweight machine learning models to determine the dubbing quality (e.g., a dubbing quality score) of a real-time generated dubbed segment and identify the cause of low-quality dubbing segments (e.g., the root cause of a low-quality score). In addition, the video dubbing system provides proactive indications to a video player to signal poor-quality dubbing segments before or while they play. Furthermore, the video dubbing system can provide reasoning behind why a particular segment of a streaming video has low-quality dubbing before or when a dubbed audio segment begins playback.
G10L 25/60 - Speech or voice analysis techniques not restricted to a single one of groups specially adapted for particular use for comparison or discrimination for measuring the quality of voice signals
G06F 40/58 - Use of machine translation, e.g. for multi-lingual retrieval, for server-side translation for client devices or for real-time translation
G10L 15/06 - Creation of reference templatesTraining of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
G10L 25/57 - Speech or voice analysis techniques not restricted to a single one of groups specially adapted for particular use for comparison or discrimination for processing of video signals
G11B 27/031 - Electronic editing of digitised analogue information signals, e.g. audio or video signals
Systems, methods, and computer program products are disclosed for generating semantic hashes using a language model (LM). A semantic hash is generated for an input by determining a plurality of strings from the input, combining the plurality of strings to generate input text, and chunking the input text into a plurality of chunks based on an input limit of the LM. Chunk embeddings are determined for the plurality of chunks using the LM, and combined to generate the semantic hash.
An augmentation service receives a request from a client for a set of actions to suggest with respect to content encoded in an image file included with the request. The service sends a first request to a content generation service to obtain the set of actions, including a first prompt that tasks a small language model (SLM) to generate the set of actions based on the content in the image file. The augmentation service replies to the client with at least a portion of the set of actions. The augmentation service receives, from the client, an indication of a selected action of at least the portion of the set of actions and sends a second request to the content generation service to perform the selected action. The second request includes a second prompt that tasks a large language model (LLM) to perform the selected action.
G06V 10/82 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
G06V 10/764 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
G06V 10/774 - Generating sets of training patternsBootstrap methods, e.g. bagging or boosting
60.
APPLICATIONS OF RETRIEVAL-AUGMENTED GENERATION FOR SOFTWARE CODE
This document relates to processing of source code using generative language models. One example method includes accessing source code, processing the source code to identify entities in the source code, and generating a graph having nodes representing the entities in the source code and edges representing relationships among the entities. The example method also includes prompting a generative language model to generate augmentation data for the entities based at least on the relationships, receiving, from the generative language model, generated augmentation data, and generating an augmented graph by associating the generated augmentation data with respective nodes of the graph. The augmented graph provides a basis for subsequent operations on the source code by the generative language model.
Techniques are described herein that are capable of increasing efficiency of a kernel using streaming multiprocessor-level time estimation. Tiling strategies for performing respective implementations of a matrix multiplication operation are defined by taking into consideration dimensions of first and second matrices that the matrix multiplication operation is configured to multiply. Estimated amounts of time or estimated latencies, which are associated with a kernel performing the respective implementations of the matrix multiplication operation using the respective tiling strategies, are calculated. The kernel is configured to implement an identified tiling strategy to perform a subsequent implementation of the matrix multiplication operation as a result of an estimated amount of time associated with the identified tiling strategy being no greater than an estimated amount of time associated with each other tiling strategy.
Dynamically loading endpoint data during System-on-Chip (SoC) validation in processor-based devices is disclosed herein. In one exemplary embodiment, a processor-based device, by executing an SoC validator, obtains a path for an endpoint node of an endpoint tree data structure. The path comprises node identifiers including a root node identifier of a root node, intermediate node identifiers of corresponding intermediate nodes, and an endpoint node identifier of the endpoint node, and corresponds to a hierarchical path from the root node of the endpoint tree data structure to the endpoint node. The SoC validator traverses the endpoint tree data structure from the root node to the endpoint node based on the path, and retrieves value data for the endpoint node based on the traversal. The SoC validator then generates an endpoint object representing the endpoint using the value data, and performs an access operation on an endpoint using the endpoint object.
G06F 21/71 - Protecting specific internal or peripheral components, in which the protection of a component leads to protection of the entire computer to assure secure computing or processing of information
G06F 15/78 - Architectures of general purpose stored program computers comprising a single central processing unit
63.
PHASE-LOCKED LOOPS (PLL), INCLUDING TIME-TO-DIGITAL CONVERTER (TDC) GAIN CALIBRATION CIRCUITS AND RELATED METHODS
In a calibrated phase-locked loop (PLL), a time-to-digital (TDC) converter circuit can be calibrated to a nominal gain by a calibration circuit to achieve a desired jitter response in the PLL. The TDC circuit in the PLL measures a time difference between the reference clock and a feedback signal as a number of time increments, and the calibration circuit adjusts a resolution of the measurement by adjusting the length of the time increments (i.e., resolution). In a Vernier method employed to measure the time difference, the length of a time increment is determined by a delay difference between a first delay of a first delay circuit in a first series of first delay circuits and a second delay of a second delay circuit in a second series of second delay circuits. Adjusting the resolution of the TDC circuit includes adjusting the delay difference between the first delay and the second delay.
Processor-based system supporting in-field testing using external dynamic random access memory (DRAM) for storing and accessing test scan data. The processor-based system includes a processor that includes one or more central processing units (CPUs) that each have access to resources, such as cache memory, a memory controller to access system memory (e.g., DRAM), interfaces circuits, to perform tasks by executing of program code. The processing-based system includes an internal, built-in testing system that allows the processor-based system to be placed into test mode to perform in-field testing of the processor-based system. To support larger-sized scan data, the processor-based system is configured for the built-in-test system to access test scan data stored in DRAM in the processor-based system in a test mode. In this manner, the DRAM supports storing larger-sized test scan data so that greater in-field test coverage can be performed in the processor-based system.
Various embodiments discussed herein are directed to improving existing technologies by providing a corpus data supplement as input into a model, such as a Large Language Model (LLM). Consequently, the model can generate accurate scores or data for predictions because the model is better able to distinguish between a general understanding of natural language concepts and domain-specific concepts.
The disclosure herein describes training a document recommendation model using loss data generated from a linear score difference vector. Training data is serialized, the training data comprising training data entries, a training data entry comprising a query, candidate documents, and labels corresponding to the candidate documents, wherein serializing the training data avoids truncating or padding the candidate documents. The training data is provided to a document recommendation model. Document prediction scores are obtained from the document recommendation model, the document prediction scores indicative of a likelihood that the candidate documents are responses to the query.
A method (200) for dynamic prompt generation includes receiving document content data (116) representing a digital document (102) accessed during a document access session (104). One or more document context parameters (120) are received relating to a context of the document access session (104). A current prompt domain (122) is determined that pertains to the document access session (104). From a prompt generation system (112), one or more candidate prompts (124) are received, specifying one or more respective machine learning (ML)-mediated document interaction operations (128) that could be applied to the digital document (102). The one or more candidate prompts (124) are generated based at least in part on the document content data (116), the one or more document context parameters (120), and the current prompt domain (122). The one or more candidate prompts (124) are displayed in a user interface (UI) (106). A user selection of a selected prompt (308A) is received, and the ML-mediated document interaction operations (128) associated with the selected prompt (308A) are applied.
A method implemented in an artificial intelligence (AI) offload die within a system-in-a-package involves hybrid inferencing of an AI model by a remote computing system and a compute die in the system-in-a-package. The method includes identifying a portion of the AI model for use by the compute die, utilizing a network controller in the AI offload die to fetch this portion from the remote computing system, and communicating it to the compute die. Additionally, the network controller in the AI offload die synchronizes AI model inferencing state between the compute die and the remote computing system, ensuring coordinated hybrid AI model inferencing. This approach facilitates efficient distribution and execution of AI tasks between the compute die and the remote computing system, enhancing computational performance and resource utilization.
This disclosure describes a framework for analyzing dubbed audio segments (audio translations converted into translated speech) of videos where the dubbed audio segments are generated in real time, including being generated locally on a client device. For instance, this disclosure describes a video dubbing system that utilizes various lightweight machine learning models to determine the dubbing quality (e.g., a dubbing quality score) of a real-time generated dubbed segment and identify the cause of low-quality dubbing segments (e.g., the root cause of a low-quality score). In addition, the video dubbing system provides proactive indications to a video player to signal poor-quality dubbing segments before or while they play. Furthermore, the video dubbing system can provide reasoning behind why a particular segment of a streaming video has low-quality dubbing before or when a dubbed audio segment begins playback.
G10L 25/57 - Speech or voice analysis techniques not restricted to a single one of groups specially adapted for particular use for comparison or discrimination for processing of video signals
09 - Scientific and electric apparatus and instruments
Goods & Services
Computers; Tablet computers; Laptop computers; Mobile computers; Reader for e-books and other electronic publications; Computer peripherals; Wireless computer peripherals; Computer keyboards; Computer mouse; Digital pens; Power cords; Battery chargers; Electrical cables and cord sets; Electronic docking stations; Computer docking stations; Adapters for use with computers and computer peripherals; USB hardware devices; Carrying cases and holders for electronic equipment, namely, tablet computers, laptop computers, mobile computers, readers for e-books and other electronic publications
71.
Display screen with animated graphical user interface for user connection and engagement
Microscopic light emitting diodes (micro-LEDs) systems having in situ current measurement circuits are described. An example micro-LED system includes a set of micro-LEDs formed in a display substrate and a set of pixel driver circuits formed in a backplane substrate, coupled to the display substrate, where a respective pixel driver circuit is to provide current to a respective micro-LED during a first mode of operation for the set of micro-LEDs. The micro-LED system further includes a current measurement circuit, formed in the backplane substrate, comprising an operational amplifier configured to drive a source-follower transistor. The micro-LED system further includes a set of pass transistors to, on a per pixel driver circuit basis, selectively redirect current from one or more of the set of pixel driver circuits to the current measurement circuit allowing for in situ measurement of the redirected current within the backplane substrate.
Described are examples for controlling a dimming panel for a display device. A first indication of a measured amount of ambient light can be received. A second indication of a location at which an eye is gazing can be received. A power control signal can be transmitted to a section of the dimming panel that corresponds to the location to facilitate activating dimming in the section of the dimming panel.
A system processes encapsulated packets by computing a hash from the inner packet's IP header values, selecting, based on the hash, an appliance IP address (AIPA) from a group of AIPAs, and replacing the outer packet's destination IP address with the selected AIPA before forwarding the packet to the corresponding network appliance. In another aspect, a method for applying a user-defined policy involves receiving the policy which identifies traffic subject to the policy and a rule to be enforced on the traffic. Based on the policy, a group of AIPAs is determined and appliances associated with the group of AIPAs are configured to enforce the policy. The first hop switch is then configured with the AIPAs, enabling it to select an AIPA address based on a hash and replace the outer destination IP address of the traffic with the selected virtual IP address.
H04L 9/06 - Arrangements for secret or secure communicationsNetwork security protocols the encryption apparatus using shift registers or memories for blockwise coding, e.g. D.E.S. systems
H04L 45/64 - Routing or path finding of packets in data switching networks using an overlay routing layer
Described are examples for generating a stack of images for a gallery view in an image viewing application. Multiple slices of multiple images can be identified based on a timestamp associated with each of the multiple images. Within a given slice of the multiple slices of images, an embedding can be generated for each image. A portion of the multiple images within the given slice can be grouped into the stack of images based on comparing respective embeddings for the portion of images. A single top image representing the stack of images can be displayed in the gallery view.
A database receives a query from a client for a database that has two segments. The query requires documents in the database having terms satisfying a criterion. Upon receiving the query, the system accesses a first inverted index in the first segment to find documents with terms that satisfy the expression. It identifies within the first segment first and second documents with first and second terms, respectively. Then, it accesses a second inverted index in the second segment and identifies an indication that the second term was removed from the second document. The database compiles a list of document identifiers, including the first document's identifier. The second document's identifier is excluded based on the removal indication. The list of document identifiers is used in generating a query response.
A medium voltage, modular rack system includes: an IT (Information Technology) rack configured to house a plurality of compute resources; a power panel rack comprising a plurality of power connectors and configured to operatively couple the compute resources to a low voltage DC power source; an energy storage rack comprising one or more energy storage devices, wherein each of the energy storage devices is configured to store energy from and provide energy to the low voltage DC power source; a power conversion rack comprising power conversion devices configured to convert a medium voltage power source to the low voltage DC source; and a medium voltage rack configured to receive a medium voltage AC source and a medium voltage DC source.
The described technology provides a device including. a transmitter configured on a transmitting die, a receiver configured on a receiving die, an interposer configured to communicate between the transmitting die and the receiving die at a test frequency that is equal or lower than the normal operating frequency of the system on a system on chip (SoC), and a plurality of pins configured on each of the transmitting die and the receiving die, wherein the plurality of pins are configured to communicate with the interposer at the test frequency.
A system may extract first input entities corresponding to a knowledge domain from input data and first output entities corresponding to the knowledge domain from output data, wherein a set of first entities includes one or more of the first input entities or the first output entities. The system may map at least some of the first input entities to at least some of the first output entities. The system may output the set of first entities indicating a mapping status for each of the set of first entities.
A method for dynamic prompt generation includes receiving document content data representing a digital document accessed during a document access session. One or more document context parameters are received relating to a context of the document access session. A current prompt domain is determined that pertains to the document access session. From a prompt generation system, one or more candidate prompts are received, specifying one or more respective machine learning (ML)-mediated document interaction operations that could be applied to the digital document. The one or more candidate prompts are generated based at least in part on the document content data, the one or more document context parameters, and the current prompt domain. The one or more candidate prompts are displayed in a user interface (UI). A user selection of a selected prompt is received, and the ML-mediated document interaction operations associated with the selected prompt are applied.
G06F 16/9535 - Search customisation based on user profiles and personalisation
G06F 3/04842 - Selection of displayed objects or displayed text elements
G06F 3/04845 - Interaction techniques based on graphical user interfaces [GUI] for the control of specific functions or operations, e.g. selecting or manipulating an object, an image or a displayed text element, setting a parameter value or selecting a range for image manipulation, e.g. dragging, rotation, expansion or change of colour
Technology is disclosed for adapting a dynamic audio stream to modify sounds of a sound class based on user preferences in systems including gaming systems. A user may provide, via a graphical user interface, at least one sample instance of a sound class and a transformation type. The system may use the instances of the sound class to train a generative artificial intelligence (AI) model to identify instances of the sound class in a dynamic audio stream (e.g., from game instances on a gaming instance) and to transform the instances of the sound class with the transformation type in the dynamic audio stream. The output stream from the generative AI model can be used as the audio output. As the user hears other instances of the sound class in the output audio stream, the user can provide feedback to tune the generative AI model for the user-specific instances.
A63F 13/54 - Controlling the output signals based on the game progress involving acoustic signals, e.g. for simulating revolutions per minute [RPM] dependent engine sounds in a driving game or reverberation against a virtual wall
Techniques for assisting visually impaired individuals when working with images are disclosed. A service accesses an image of a scene. The image includes pixels representing an object included in the scene. The service generates a classification for the object pixels and a classification for the scene. The service receives user input directed to the image. The user input includes at least one of: a cursor hovering over one or more of the object pixels, a selection of the one or more object pixels, or a movement of the cursor over the one or more object pixels. In response to the user input, the service triggers playback of an audio output comprising audio details describing the object classification.
G06F 3/0481 - Interaction techniques based on graphical user interfaces [GUI] based on specific properties of the displayed interaction object or a metaphor-based environment, e.g. interaction with desktop elements like windows or icons, or assisted by a cursor's changing behaviour or appearance
G06V 10/764 - Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
G09B 21/00 - Teaching, or communicating with, the blind, deaf or mute
A leak mitigation system for a fluid-cooled computing device includes an absorbent pad having a fluid sensor connected to an absorbent material. The absorbent pad is placed in a position relative to the fluid-cooled computing device to collect fluid from the cooling system. The sensor detects the presence of a fluid absorbed by the absorbent material and the leak mitigation system implements a leak mitigation protocol to prevent or reduce damage to the computing device.
B01D 12/00 - Displacing liquid, e.g. from wet solids or from dispersions of liquids or from solids in liquids, by means of another liquid
D02G 3/04 - Blended or other yarns or threads containing components made from different materials
D03D 15/283 - Woven fabrics characterised by the material, structure or properties of the fibres, filaments, yarns, threads or other warp or weft elements used characterised by the material of the fibres or filaments constituting the yarns or threads synthetic polymer-based, e.g. polyamide or polyester fibres
84.
INDIVIDUAL POWER CYCLE CONTROL OF ACCELERATOR MODULES CONFIGURED ON A NODE
Disclosed herein is a system for implementing a management controller on a node, or network server, that is dedicated to monitoring the individual health of a plurality of accelerator modules configured on the node. Based on the monitored health, the management controller is configured to implement autonomous power cycle control of individual accelerator modules. The autonomous power cycle control is implemented without violating the requirements of standards established for accelerator modules (e.g., OPEN COMPUTE PROJECT requirements, PERIPHERAL COMPONENT INTERCONNECT EXPRESS (PCIe) interface requirements).
An integrated circuit and method are disclosed for processing of compressed columnar data. The processing of compressed columnar data includes loading compressed columnar data associated with a first and a second column into a first on-chip buffer, performing time-shared processing of the compressed columnar data from the first on-chip buffer into a second on-chip buffer, transcoding the compressed columnar data from the second on-chip buffer into unified columnar data having a unified format, loading the unified columnar data into a third on-chip buffer so that the unified columnar data are logically aligned in the third on-chip buffer, and providing at least a portion of the unified columnar data to a query operator. The integrated circuit includes a column loader, a balancer, a transcoder, and a decoder for performing processing of the compressed columnar data.
A method for controlling user access to a touch-screen system (14) comprises: (a) receiving an uplink signal from a key device (18) and extracting corresponding uplink data from the uplink signal; (b) providing challenge data in response to the uplink data and transmitting a corresponding challenge signal to the key device (18), where the challenge signal is transmitted by a touch-sensor transmitter (36) also configured to transmit a synchronization signal to an active pen (46); (c) receiving a downlink signal from the key device (18) and extracting corresponding downlink data from the downlink signal, where the uplink and downlink signals are received by a touch-sensor receiver (38) also configured to receive sensory signal from the active pen (46); and (d) forbidding a user (12) from accessing the touch-screen system (14) unless the downlink data authenticates the user (12) and signal of pre-determined signal strength continues to be received from the key device (18).
One example provides a computing device (118) comprising a regular expression (regex) hardware accelerator (120) including a deterministic finite automaton (DFA) engine (126) configured to execute an object file (128), and a compiler (132). The compiler (132) is executable to generate the object file (128) based at least upon a target DFA graph by receiving a predicate including one or more of an integer condition or a floating-point condition (902), transforming the predicate to form a rewritten predicate with an equivalent expression (904), and building the target DFA graph based at least upon the rewritten predicate (910).
Robotic systems (10) and methods (200) for operating a robotic system (10) to perform inspection and cleaning of optical fiber components are disclosed. A robotic system (10) for performing automated inspection and cleaning of optical fiber components comprises a transceiver receptacle (14) moveably secured to a chassis (18) and configured to removably retain an optical fiber transceiver (20). A clamp (40) moveably secured to the chassis (18) is configured to remove an optical fiber connector (34) from the transceiver (20) and reinsert the optical fiber connector (34) into the transceiver (20). An inspection tool (22) is non-moveably affixed to the chassis (18) and configured to inspect one or more fiber ends (21, 23) of the transceiver (20) and one or more fiber ends (51, 53) of the optical fiber connector (34). A cleaning tool (26, 30) is non-moveably affixed to the chassis (18) and configured to clean the fiber end(s) (51, 53) of the optical fiber connector and the fiber end(s) (21, 23) of the optical fiber transceiver (20).
A method for processing a glass preform for hollow core fiber comprises providing a length of glass preform for hollow core fiber from which a portion is to be removed in order to terminate the preform forming an end face. The preform comprises a transverse cross-sectional structure comprising a hollow core surrounded by a plurality of capillaries defining a plurality of voids encased by a jacket tube, wherein the hollow core and the plurality of voids extend longitudinally along the length of the preform. The method further comprises positioning the preform using a preform holder attached to a diamond wire saw, wherein the preform holder is configured to clamp the preform equally on both sides of a desired location for cutting. The preform is cut in the desired location using the diamond wire saw.
B26D 1/46 - Cutting through work characterised by the nature or movement of the cutting memberApparatus or machines thereforCutting members therefor involving a cutting member which does not travel with the work having an endless band-knife or the like
B23D 57/00 - Sawing machines or sawing devices not covered by one of groups
B26D 1/547 - Cutting through work characterised by the nature or movement of the cutting memberApparatus or machines thereforCutting members therefor involving a cutting member which does not travel with the work having a wire-like cutting member
B28D 1/08 - Working stone or stone-like materials, e.g. brick, concrete, not provided for elsewhereMachines, devices, tools therefor by sawing with saw blades of endless cutter-type, e.g. chain saws, strap saws
B28D 5/04 - Fine working of gems, jewels, crystals, e.g. of semiconductor materialApparatus therefor by tools other than of rotary type, e.g. reciprocating tools
Methods and apparatuses for improving the yield and performance of integrated circuit structures by utilizing rotatable chiplets are described. During manufacturing of an integrated circuit structure that includes multiple chiplets arranged within a plurality of integration layers, each integration layer may be dynamically rotated or oriented prior to being bonded based on chiplet characteristics of the chiplets within the integration layers. Each integration layer comprises one or more chiplets. An automated manufacturing system determines the degree of rotation of a first integration layer relative to a second integration layer to which the first integration layer is to be directly or indirectly attached based on chiplet performance, capacity, and/or thermal characteristics of the chiplets within the integration layers.
A method for processing an image of a hollow core fiber, HCF, is described. For an edge of a tube of the HCF, brightness of the image is used to detect points corresponding to the edge. The method further includes fitting an edge model function to detected edge points, and identifying an outlier point of the detected edge points that is above a threshold distance from the fitted model. The outlier point is removed and the model is refitted to the remaining points. The method comprises iteratively identifying and removing subsequent outlier points and refitting the model to remaining points until all remaining points are inlier points below a final distance threshold from the model. Remaining inlier points are fitted to a final model. The final model is used to determine a geometric parameter of the HCF for use during quality control, splicing and/or fiber drawing.
The disclosure relates to power distribution systems for Information Technology (IT) racks. For example, a power distribution system may include a sidecar unit that provides flexibility for enhanced power delivery, including High Voltage Direct Current (HVDC) power of at least 400V or more to meet the power demands of modern data centers and other computer facilities. The sidecar unit may also provide backward compatibility with IT racks that use 50V inputs for its components. The power distribution system may include an architecture in which an IT rack includes rack slots that each house a node and a corresponding sidecar unit, which can be released from the node for easy serviceability of the sidecar unit and/or the node. The IT rack may take HVDC power as input from a power rack having power distribution units.
Systems, methods, and computer program products are described herein related to on-demand battery life with adaptive performance control, which achieves a target battery life (e.g., based on user request) by dynamically regulating a battery discharge slope. A power mode is implemented with adaptive power level (PL) limits (e.g., average PL limit and maximum PL limit) to manage battery life over time intervals. The discharge rate of the battery is controlled by dynamically adapting PL limits of the power provided to the computing device to align the remaining battery life with a target battery life. PL limits can be adapted based on a hardcoded mapping of a target performance levels or based on a machine learning model. Computing device performance is adaptively enhanced on demand for usage scenarios by temporary adjustment of PL limits when estimated battery life is within a margin of target battery life.
Systems and methods for automatically reducing regression for a software update applied to a population of nodes in a computing environment. A regression detector performs a health analysis of the software update and detects a software regression attributed to the software update with high confidence by performing a combination of data analyses. In some examples, a time window-based observational study, a control-based observational study, and an anomaly detection analysis are performed for identifying various regression conditions. When the identified regression conditions match a set of high-confidence regression conditions configured for the health analysis, a software regression is detected. In further examples, the regression detector transmits an event based on the detected software regression to prevent the software regression from propagating to additional nodes in the computing environment.
A mobile computing device is configured to charge a component battery. The mobile computing device comprises a power supply unit, a computing device battery, a processor, and a memory storing instructions executable by the processor to control a rate of charging the component battery from either the power supply unit or the computing device battery based at least in part on a magnitude of throttled power provided to the processor.
Robotic systems and methods for operating a robotic system to perform inspection and cleaning of optical fiber components are disclosed. A robotic system for performing automated inspection and cleaning of optical fiber components comprises a transceiver receptacle moveably secured to a chassis and configured to removably retain an optical fiber transceiver. A clamp moveably secured to the chassis is configured to remove an optical fiber connector from the transceiver and reinsert the optical fiber connector into the transceiver. An inspection tool is non-moveably affixed to the chassis and configured to inspect one or more fiber ends of the transceiver and one or more fiber ends of the optical fiber connector. A cleaning tool is non-moveably affixed to the chassis and configured to clean the fiber end(s) of the optical fiber connector and the fiber end(s) of the optical fiber transceiver.
Described are examples for controlling a chromatically adaptive film for a display device. An indication of one or more colors for absorption by the chromatically adaptive film can be received. Based on the indication, light transmittance in one or more color absorptive layers of multiple color absorptive layers of the chromatically adaptive film can be modified.
G09G 3/20 - Control arrangements or circuits, of interest only in connection with visual indicators other than cathode-ray tubes for presentation of an assembly of a number of characters, e.g. a page, by composing the assembly by combination of individual elements arranged in a matrix
Techniques are described herein that are capable of deploying a branch of a main codebase based on compliance of the branch with a security policy since creation of the branch(es). A first branch of a main codebase is stored in a designated store and a second branch of the main codebase is not stored in the designated store as a result of the first branch complying with a security policy throughout a first time period since creation of the first branch and the second branch failing to comply with the security policy during a second time period since creation of the second branch. As a result of the first branch being stored in the designated store and complying with the security policy throughout the first time period, the first branch is converted into a deployable code branch, and the deployable code branch is deployed.
Technology is disclosed for adapting a dynamic audio stream to modify sounds of a sound class based on user preferences in systems including gaming systems. A user may provide, via a graphical user interface, a selection of a sound class and a transformation type. Using a pretrained generative artificial intelligence (AI) model, the system may process the dynamic audio stream (e.g., for a gaming instance) to transform instances of the sound class with the transformation type. The output stream from the generative AI model can be used as the audio output. The described technology allows for processing and transforming all dynamic audio streams on a system without specific programming for a particular application or game.
A63F 13/54 - Controlling the output signals based on the game progress involving acoustic signals, e.g. for simulating revolutions per minute [RPM] dependent engine sounds in a driving game or reverberation against a virtual wall
