Abstract: Systems and methods for a user interface for preventing data transmission including: applying data processing rules to a set of transaction data to determine a furnishing decision for a subset of the set of transaction data, receiving a selection of one or more data processing rules of the set of data processing rules; determining, based on the selected data processing rules, a list of accounts comprising at least a first account and a second account, wherein the first account is associated with the first data of the subset and the furnishing decision for the first data, and the second account is associated with the second data of the subset and the furnishing decision for the second data; changing the furnishing decision for the first data; and automatically updating the furnishing decision for data associated with each account of the list of accounts.

Abstract: Systems and techniques are provided for processing image data. According to some aspects, a computing device can generate a gradient (e.g., a classifier gradient using a trained classifier) associated with a current sample. The computing device can combine the gradient with an iterative model estimated score function or data associated with the current sample to generate a score function estimate. The computing device can predict, using the diffusion machine learning model and based on the score function estimate, a new sample.

Abstract: A method of operating a server is provided that includes providing, with the server, one or more services relating to network access control and management of a network, predicting a failure of the server with a failure prediction model, and performing a remedial action to avoid disruption in the one or more services. The server can be a network access control and management server. The failure prediction model can be a machine-learning based server failure prediction model that is trained on past server failure events. Operated in this way, the server can operate with minimal network/server disruption so that service level agreements with various customers are maintained while reducing operational complexity.

Abstract: Various embodiments of the present disclosure provide machine learning techniques for transforming disparate, third-party datasets to canonical representations. The techniques include generating, using a machine learning prediction model, a canonical representation for an input dataset. The machine learning prediction model is previously trained using permutative input embeddings for a training dataset based on canonical data entity features, such that each permutative input embedding corresponds to a different sequence of the canonical data entity features. The permutative input embeddings are leveraged to generate a latent representation for the training dataset. The latent representation is combined with a canonical data map to generate an alignment vector, which is refined to generate an output vector for the input dataset. The machine learning prediction model is trained using a model loss generated based on a comparison of the output vector with a corresponding labeled vector.

Abstract: An integrated circuit includes memory circuits, a selector circuit, a bus coupled to the selector circuit, and a controller circuit. The controller circuit provides test signals from the controller circuit through the bus to the selector circuit for transmission to the memory circuits during a memory built-in-self-test mode. Each of the memory circuits can include a comparator circuit configurable to compare a read data bit read from one of the memory circuits to an expected data bit in the test signals to generate a sticky error bit during the memory built-in-self-test mode.

Abstract: According to one or more embodiments of the present disclosure, a semiconductor device is described. The semiconductor device may include a substrate, a channel portion on the substrate between a source region and a drain region, and a gate on the channel. The channel portion may include a first portion extending in a first direction and at least one second portion protruding from the first portion in a second direction crossing the first portion.

Abstract: A method and system for a content broker, including a unified object index, where the content broker is coupled to the unified object index and receives, from a requesting entity, a request to perform an action on an object and the object is stored in the content repository. The method further including obtaining the object associated with the request from a content repository, determining, using the unified object index, a normalized object type associated with the object, obtaining a governance rule based on the normalized object type, and servicing the request using the governance rule.

Abstract: Described herein are novel stable 9-N-, 8-N-, and 7-N-acetyl analogues of instable 9-O-acetyl, 8-O-acetyl, and 7-O-acetyl b-series gangliosides and glycosphingosines, including GD3, GD2, GD1b, GT1b, GQ1b, and their glycosphingosines. Chemoenzymatic methods for the production of the stable 9-N-, 8-N-, and 7-N-acetyl analogues are also described herein.

Abstract: Impeller architecture for a cooling fan and methodology for making same. The impeller architecture includes a plurality of blades, individual ones of the blades have a first end that is attached to a hub component in a sequential order, such that sequential first ends are attached to the circumference. An indexing function is applied to the sequential order, and blades or the spaces therebetween are modified accordingly to have a blade type based on their sequential location and the indexing function. The indexing function can be, in a non-limiting example, odd numbers or prime numbers.

Abstract: Embodiments described herein provide a training framework for generative NLP models. Specifically, the training input, e.g., in the form of a sequence of tokens representing a user-agent dialogue, may be randomly masked for a few spans, which can be one or more tokens, one or more words, one or more sentences, or one or more paragraphs. These masked spans are replaced with their embeddings generated from pre-trained large language models are then used for training the NLP model.

Abstract: A method of operating a network is provided that includes receiving a query, using a first language model to determine an intent or purpose of the query, using a second language model to extract a named entity from the query, and obtaining search results by searching for the extracted named entity on a named entity list corresponding to a particular tenant. The method can further include generating a response based on the search results. The query can be a natural language query, and the first language model can be a natural language model. The second language model for extracting the named entity can be a network-related language model that is trained on network records associated with a plurality of tenants. The network records associated with the plurality of tenants can be stored on a multi-tenant database.

Abstract: A memory system to store multiple bits of data in a memory cell. After receiving the data bits, a memory device coarsely programs a threshold voltage of the memory cell to a first level representative of a combination of bit values according to a mapping between combinations of bit values and threshold levels. The threshold levels are partitioned into a plurality of groups, each containing a subset of the threshold levels. A group identification of a first group, among the plurality of groups, containing the first level is determined for the memory cell. The memory device reads, using the group identification, a subset of the data bits back from the first memory cell, and combines the bits of the group identification and the subset to recover the entire set of data bits to finely program the threshold voltage of the memory cell to represent the data bits.

Abstract: A stepped diffraction grating and method for manufacturer thereof are disclosed. A plurality of parallel grating lines are each formed on a substrate surface by forming a plurality of stacked layers of optically transmissive material. In cross-section, each grating line has an upper layer having an upper surface having a first end and a second end; a bottom layer having a bottom surface abutting the substrate surface and an upper surface having a first end and a second end; a rising staircase portion extending at a rising staircase angle between 10 degrees and 60 degrees; and a falling staircase portion extending at a falling staircase angle between the rising staircase angle and 89 degrees.

Abstract: A device access management server may facilitate secure access of a target device by an accessing device. The secure remote access of the target device by the accessing device may be facilitated by a public key infrastructure (PKI) certificate issued and/or validated by the device access management server.

Abstract: IoT service layer capabilities may be employed to automate and simplify the service enrollment process for IoT service subscribers/enrollees. These capabilities enable virtualization of a service subscriber and the physical IoT devices, applications, data and authorized users of the subscriber into a software profile that is representative of the subscriber. Once virtualized, a service subscriber may then delegate the complexities and burden of service enrollment to an automated IoT service enrollment software function.

Abstract: This disclosure provides systems, methods, and devices for vehicle driving assistance systems that support image processing. In a first aspect, a method of image processing includes receiving an image frame from an image sensor of a camera; receiving an indicator associated with a type of lens of the camera; determining a first tensor grid associated with the indicator, the first tensor grid including a plurality of image framework positions associated with the type of lens; and determining, using a machine learning model, a BEV feature map corresponding to the image frame based on features of the image frame and the first tensor grid. Other aspects and features are also claimed and described.

Abstract: Systems and methods related to improved program guide are provided. In one example, a method includes: receiving a first user request for browsing a program guide for digital programs on a client device, transmitting data corresponding to the program guide to the client device, and displaying the program guide in a user interface. The program guide includes multiple program preview boards, and each program preview board corresponds to a digital program or a recorded digital program and includes information related to the digital program or the recorded digital program. The method further includes providing a first selectable option for the user to enable or disable a first spoiler prevention feature and enabling or disabling the first spoiler prevention feature based on the user's selection. The first spoiler prevention feature is configured to hide a part of the information related to the digital program in the corresponding program preview.

Abstract: An application client on a user device obtains an authorization token from the computing platform that hosts one or more edge application devices, initiates an edge discovery request to the computing platform using the authorization token, and, in response to the edge discovery request, receives connection information of a particular edge application device among the one or more edge application devices. The application client establishes a connection between the application client and the particular edge application device using the connection information.

Abstract: Disclosed herein are approaches for in-situ verification and correction of a wafer position. In one approach, a method may include illuminating an underside of a platen positioned within a processing chamber, and detecting a perimeter edge of the platen using an imaging device positioned external to the processing chamber, above the platen. The method may further include determining, via a controller, position data for the platen based on the detected perimeter edge of the platen, and positioning a wafer atop the platen based on the position data of the platen, wherein the wafer comprises a positioning notch.

Abstract: Embodiments described herein provide a training framework for generative NLP models that operate on previously learnt knowledge from pretrained large language models. Specifically, to train an NLP model to generate a response to a user utterance (e.g., “resolve login issue”), document embeddings of support IT documents encoded by a pretrained LLM are fed to an NLP decoder together with a training dialogue (e.g., a dialogue between the chat agent on how to “resolve login issue”). The NLP decoder can thus be trained by a causal language modeling loss computed based on the predicted next token and the ground-truth token from the training dialogue.