Abstract: A method includes receiving a machine learning (ML) network model in high-level code; generating an internal representation (IR), the IR is mapped to components in a multi-processing tile device; determining whether a first processing tile with a first on-chip memory (OCM) has a same dimension for an input/output tensor data as a second processing tile with a second OCM performing a same primitive function based on the IR; allocating a same memory address range within the first and the second OCM for the same primitive function if the first processing tile has the same dimension for the input/output tensor data as the second processing tile for the same primitive function; linking the memory address range of the first OCM to the memory address range of the second OCM to form a grouped memory space within the first and the second OCM respectively; and compiling low-level instructions based on the linking.

Abstract: A system and method provide access by a web application running on a host computing device in communication with a remote server to a native binary DLL. A browser extension to the web application and a native messaging application communicate with the browser extension via standard input/output. A long-running computational process on the host computing device configured as a wrapper for the binary DLL is executable by the host computing device. The native messaging application and the long-running computational process communicate with one another via a named pipe, and the long-running computational process duration is independent of the native messaging application duration.

Abstract: System, method, and circuitry for generating content for a programmable computing device based on user-selected configuration information. A settings registry is generated based on the user's selections. The settings registry and the user selected configuration information is utilized to generate the content, such as code, data, parameters, settings, etc. When the content is provided to the programmable computing device, the content initializes, configures, or controls one or more software and hardware aspects of the programmable computing device, such as boot sequence configurations, internal peripheral configurations, states of the programmable computing device, transitions between states of the programmable computing device, etc., and various combinations thereof.

Abstract: Some embodiments generate prompts and submit them in queries to a language model trained on code to perform automated program repair. Some embodiments fix syntactic mistakes and semantic mistakes by combining multimodal prompts, iterative querying, test-case-based selection of few-shots, and program chunking. In some cases, edit distance is minimized between an initial flawed program and the automatically created improved version of that program. The initial flawed program is obtained from a programming student, or from a source code generator.

Abstract: An elidable text is constructed that prioritizes the content included in a prompt to a large language model having a fixed-size context window. The elidable text is generated from developer-generated instructions or automatically for source code within a source code editor. A source code editor may include a feature that selects certain lines of code as important or focused which are assigned a high-priority value. A changed line, a line of source code at a current cursor position, lines of source code at the beginning of a file and those that output data are considered focused lines. Non-focused lines are assigned a priority based on a distance from a focused line. The elidable text constrains the data included in a prompt to the context window size by replacing the lowest-valued lines of text and source code with a replacement string.

Abstract: An application installation method and apparatus, a terminal device, and a storage medium are provided, and are applicable to the terminal control field. The method includes: a first device obtains installation information of a first application from a second device when the first device detects a specified operation on the first application in a current display interface, where the first application is an application on a home screen of the second device, and the home screen of the second device is displayed in the current display interface of the first device; and the first device installs the first application based on the installation information of the first application. The method can conveniently implement application sharing between devices in a multi-screen collaborative operation process.

Abstract: A memory operation device and method for operating a memory in an electronic device. The electronic device may determine whether a memory leak occurs in one or more low-order stack trace items with a count value of n among collected stack trace items, n being a positive integer, and, based on a low-order stack trace item among the collected stack trace items being determined as causing a memory leak, proceeding to a next order of the collected stack trace items to thereby determine whether a memory leak occurs in one or more high-order stack trace items with a count value of m which is a positive integer higher than n). When m is a maximum count value among the collected stack trace items, memory debugging may be performed using a high-order stack trace item, among the one or more high-order stack trace items, causing the memory leak.

Abstract: Systems and methods for adding process actions to the design of a robotic software process. A context-recognition module recognizes a current state of a process being designed, and passes information on that current state to a recommendation module. The recommendation module evaluates the current state and identifies at least one suitable process action to recommend in response to that current state. The recommendation module then recommends the at least one process action to the human designer. If the designer accepts the recommendation, a design module adds the process action to the process design. The recommendation module may also use information about previous actions in the process and in other processes when identifying suitable process actions. The context-recognition module and the recommendation module may each comprise at least one machine learning module, which may or may not be neural network based.

Abstract: Methods and systems relating to the field of parallel computing are disclosed herein. The methods and systems disclosed include approaches for sparsity uniformity enforcement for a set of computational nodes which are used to execute a complex computation. A disclosed method includes determining a sparsity distribution in a set of operand data, and generating, using a compiler, a set of instructions for executing, using the set of operand data and a set of processing cores, a complex computation. Alternatively, the method includes altering the operand data. The method also includes distributing the set of operand data to the set of processing cores for use in executing the complex computation in accordance with the set of instructions. Either the altering is conducted to, or the compiler is programmed to, balance the sparsity distribution among the set of processing cores.

Abstract: Using a common reference address when processing calls between a native application binary interface (ABI) and a foreign ABI. Based on a caller calling using a reference address, a lookup structure is used to determine whether the reference address is within a memory range storing native code and that the callee is native, or a memory range not storing native code and that the callee is foreign. Execution of a callee is initiated based on one of (i) calling the callee using the reference address within an emulator when the caller is native and the callee is foreign; (ii) calling an entry thunk when the caller is foreign and the callee is native; (iii) calling an exit thunk when the caller is native and the callee is foreign; or (iv) directly calling the callee using the reference address when the caller is native and the callee is native.

Abstract: Methods for generating automations via natural language processing are performed by computing systems. Natural language input is received from a user interface, and an automation workflow is generated based on the natural language input. The automation workflow includes steps to build an automation. One or more of the steps is provided to the user interface, and a first field and a second field that each correspond to the one or more steps are populated in the user interface. The first field is populated with a parameter value based on the natural language input, and the second field is populated based on the parameter value. The automation is then enabled to be built and deployed.

Abstract: Aspects of the subject disclosure may include, for example, a non-transitory machine-readable medium comprising executable instructions that, when executed by a processing system including a processor, facilitate performance of operations, the operations comprising: receiving by a selected one of a plurality of containers incoming traffic from a legacy application, the legacy application having been associated with a migrated application prior to the migrated application having been installed on a Software as a Service (SaaS) cloud infrastructure, the incoming traffic comprising a first instruction to perform a first legacy operation; mapping the first legacy operation of the legacy application to a replacement instruction for the migrated application by receiving data from the selected container and assigning the replacement instruction to replace the first instruction; and communicating with the migrated application on the SaaS cloud infrastructure, the communicating comprising sending to the migrated appli

Abstract: A server is provided and is programmed to automatically determine, based on the identification data that uniquely identifies a patient device, whether there is an upgrade to the code or data (e.g., software or firmware) stored in the electronic memory of the patient device that can be applied thereto. In response to determination that there is an upgrade that can be applied, a data package is delivered to the patient device via the first wireless transceiver, the data package to be automatically applied to the electronic memory of the patient device.

Abstract: An incremental compiling method and an incremental compiling system based on a heterogeneous device. The method comprises: a user using P4 to develop a network program, checking the syntax and semantics of the network program and converting into an intermediate representation; allocating intermediate representation files according to constraints such as the functions and resources of the heterogeneous device; then, transforming the network program into different rule expressions of the heterogeneous device according to the constraints of the heterogeneous device, flow table entries and the like, and loading the on different rule expressions to a target device; when the user needs to deploy network functions incrementally, the user only needs to write new network programs and regenerate rule expressions; by comparing and analyzing the old and new rule expressions, a rule expression that can be loaded on heterogeneous device is formed, and the incremental deployment of network functions is realized.

Abstract: Computer technology for optimizing operator configuration for deployment of an application in a container orchestration cluster. The computer technology includes machine logic for performing the following operations: collecting historical data for deployment data including operator configurations, cluster configurations, and behavior data. The method models the deployment data using embedded matrix models; and applies collaborative filtering to process the modeled data to map user requirements for a target deployment to deployment configurations of previous successful deployments; to output one or more predicted deployment configurations as a reference for the target deployment.

Abstract: Automation solutions for event logging and debugging on KUBERNETES are disclosed. A method for navigating data associated with a distributed application cluster includes providing an edge module configured to deploy a Berkeley Packet Filter (BPF) probe and a corresponding BPF program in a computing environment. The BPF probe is triggered based on an event associated with a distributed application running in a user space of the computing environment, the distributed application being one entity of a plurality of entities on the distributed application cluster. Data associated with the event is collected in a kernel space of the computing environment via the BPF program. At least one relationship is identified between the distributed application and at least one entity of the plurality of entities based on the collected data. At least one recommended data set based on the at least one identified relationship is provided via a user interface.

Abstract: In one aspect, a disclosed method includes learning one or more chip agnostic parameters across a plurality of best known configuration (BKC) firmware versions, performing BKC attributes tuning based on said learning, implementing platform specific BKC table offsets and a handoff block to pass the table offsets to update routines by creating a trusted session for platform firmware table updates, and dynamically publishing changes in BKC policy. A BKC firmware serialization protocol may be implemented to ensure gaps in firmware versions at an end user platform are resolved by synchronizing each BKC version attribute. The serialization protocol may employ node-based cloud ecosystem learning. The method may further include reloading memory map parameters for uninterrupted services. The uninterrupted services may include, as examples, user presence detection after power resume and central processing unit (CPU) power cap functions.

Abstract: Techniques for implementing an integrated testing tool for data replication are disclosed. In some embodiments, a software tool performs operations comprising: causing a first set of user interface elements to be displayed on a computing device; receiving, via the first set of user interface elements, a data replication configuration identifying an adapter, a source database, and a target database; storing the data replication configuration in an embedded database; causing a second set of user interface elements to be displayed on the computing device; receiving, via the second set of user interface elements, a data generation configuration identifying a table in the source database; generating random data in the table based on the data generation configuration; instructing the adapter to replicate the source data from the source database onto the target database based on the data replication configuration; and causing a log of events to be displayed on the computing device.

Abstract: A method for providing an automated virtual assistant testing framework to test one or more components associated with a virtual assistant (VA) is disclosed. The method includes integrating, one or more VA test suites with a continuous integration and continuous deployment (CICD) pipeline, wherein the one or more VA test suites comprises test case(s); receiving in a repository, a commit command; triggering, the CICD pipeline based on the commit command; executing, the one or more VA test suites based on the triggering of the CICD pipeline; and testing, the one or more components associated with the VA using a VA skills and intent training data, wherein the one or more components comprises custom UI(s), skill(s), integration(s) and/or one or more natural language processing (NLP) capabilities, and wherein the testing is based on the execution of the one or more virtual assistant (VA) test suites.

Abstract: Removing inactive components to facilitate code generation. A first file that includes a first group of components in a first notation syntax and a second file that includes a second group of components in a second notation syntax is received. A first set of inactive components of the first group of components is identified. A second set of inactive components of the second group of components is identified. The components of the first group of components and the second group of components that are not marked as inactive components are active components. Code to be executed based on the active components is generated.