59 results about "Application domain" patented technology

Systems and methods for constructing layered prompts to operate as input into a pre-trained large language model (LLM). The method involves obtaining a set of application domains in which the LLM will be used. Using these application domains, a set of guidelines is determined, defining operation boundaries for the LLM. A set of layers is determined, each associated with the guidelines and including variables representing attributes identified within those guidelines. Using these layers, a first layered prompt is constructed to test the initial operation boundaries of the guidelines and is supplied to the LLM to generate a set of responses. Based on the responses, a second layered prompt is dynamically constructed to test additional operation boundaries, ensuring iterative refinement and contextual relevance.

A decentralized artificial intelligence platform is disclosed that enables secure sharing and execution of AI modules in sensitive application domains. The platform comprises a knowledge module registry for discovering specialized AI modules, a security framework enforcing structural integrity of AI modules, a zero-knowledge proof validation service for verifying usage conditions without exposing private data, and a blockchain-based value chain for managing smart contracts and payments. In operation, an AI agent can fetch a required AI module from the knowledge module registry, prove the right to use it, execute the AI module on local data, and automatically pay the AI module's provider through a blockchain transaction.

Systems and methods for measuring, grading, evaluating, and comparing AI models via a graphical user interface are disclosed. The technology obtains a set of application domains of the AI model in which an AI model will be used. The application domains are mapped to one or more guidelines to determine a set of guidelines that define operational boundaries of the AI model. The guidelines are used to generate assessment domains, each associated with specific benchmarks that include indicators of a degree of satisfaction with the guidelines. For each assessment domain, assessments are constructed to evaluate the AI model's degree of satisfaction with the corresponding guidelines. The AI model is then evaluated against the assessments. Based on these comparisons, grades are assigned to the AI model for each assessment domain. The application-domain-specific grades are generated and displayed at a GUI, reflecting the AI model's degree of satisfaction with the guidelines.

The invention provides a hot updating method, system and equipment for an automation program of semiconductor equipment and a medium. The method comprises the following steps: dynamically loading a new version script code into an isolated runtime environment created by an application domain or a class loader through a reflective API (Application Program Interface) so as to isolate new and old codes. And performing preheating initialization on the new code in an isolated environment, wherein the preheating initialization comprises execution of an initialization function and simulation of a production request test. After preheating is completed, a new request of the production environment is routed to the new version to be processed through atomic switching operation, and meanwhile, the current running version continues to complete existing in-process affairs. According to the method, shutdown is avoided through dynamic loading and isolation, stability is improved through preheating initialization, new requests and in-process transactions are considered in atomic switching, non-shutdown updating of the automatic program of the semiconductor equipment is achieved, production interruption is effectively eliminated, and the equipment utilization rate and operation and maintenance flexibility are improved.

The invention provides an aerospace application knowledge graph construction method, device, system and equipment and a medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: obtaining multi-modal data; based on the data types, data processing is carried out on the modal data, and lightweight markup language texts and embedded vectors are obtained; performing knowledge extraction on the lightweight markup language text on the basis of a preset cue word and a large language model to obtain metadata information, and constructing an entity relationship triple on the basis of a relationship between entities and the metadata information; and based on the aerospace knowledge file number information, determining an association relationship among the aerospace knowledge file, the lightweight markup language text, the embedded vector and the entity relationship triple, and constructing the aerospace application knowledge graph according to the association relationship and the entity relationship triple. According to the space application knowledge graph constructed by the method, the space application field entities and the association relationship can be accurately presented, and efficient and accurate knowledge support is provided for space application knowledge questions and answers.

The invention relates to an anonymous network identity authentication method and an anonymous network identity authentication system generated based on a trusted root, which realize anonymous generation and controlled privacy protection of a trusted identity identifier of an anonymous network of a user and solve the problems of leakage of user network identity information and track link. Generating a user registration private key on a user side; generating a user anonymous network trusted identity identifier corresponding to the application party based on the user registration private key; signing the trusted identity label of the anonymous network and the application domain information through a user registration private key; the application side verifies the correctness of the user signature information by using the trusted identity management platform public parameter, the application domain public parameter and the user anonymous network trusted identity identifier, and completes the anonymous network identity authentication of the user in the application domain; when there is a supervision demand for the user identity, the trusted identity management platform can trace the anonymous network trusted identity label to track the real identity information of the user.

Systems and methods for constructing layered prompts to operate as input into a pre-trained large language model (LLM). The method involves obtaining a set of application domains in which the LLM will be used. Using these application domains, a set of guidelines is determined, defining operation boundaries for the LLM. A set of layers is determined, each associated with the guidelines and including variables representing attributes identified within those guidelines. Using these layers, a first layered prompt is constructed to test the initial operation boundaries of the guidelines and is supplied to the LLM to generate a set of responses. Based on the responses, a second layered prompt is dynamically constructed to test additional operation boundaries, ensuring iterative refinement and contextual relevance.

Described is a data processing device (1901) for performing end-to-end user query correction and slot filling, the device having one or more processors configured to: receive (1801) a user query (301); process (1802) the user query to determine whether the user query is correct with respect to an application domain language or a natural language; based on the processing of the user query, form (1803) a corrected query (350); perform (1804) a slot filling operation for the corrected user query to identify one or more slots (306) for the corrected query, each identified slot corresponding to a part of the corrected user query; and classify (1805) each identified slot of the corrected query into one of a plurality of classes. This approach can handle variations of non-correct input in multilingual settings in addition to the slot filling task and can handle application domain language in addition to different natural languages.

The invention provides a feedback information generation method and device, and relates to the technical field of artificial intelligence, in particular to the technical fields of large models, automatic driving, intelligent traffic and the like. According to the specific implementation scheme, the method comprises the steps of determining an application field of current query information and a target component set associated with the application field; inputting the current query information and the target component set into a large language model, and generating a function component calling flow chart; and the execution function component calls the action instruction sequence corresponding to the flow chart to generate feedback information. According to the mode, the accuracy and efficiency of generating the feedback information are effectively improved.

A domain-specific performance / power regulation technique involves periodically identifying a domain of an application currently being executed by a given processor from among a plurality of domains of the application. The technique also includes measuring the instructions per second (IPS) of the given processor during execution of the identified domain. The method also includes determining a computational constraint parameter of the identified region based on the measured IPS. The method also includes determining an optimal frequency for the identified region based on the computational constraint parameter and instructing the given processor to set a frequency of the given processor to the optimal frequency during execution of the identified region.

The invention relates to a computer architecture for verifying image frame and image line data in a single pass. Aspects of this technical solution may increase the processing speed of low latency applications while maintaining the integrity of error correction detection at these higher speeds. For example, in an image processing environment associated with autonomous navigation (e.g., driving), a large amount of image data is to be quickly and accurately processed to maintain a reliable and latest model of the physical environment. Accordingly, embodiments according to the present disclosure may provide high-speed and accurate error correction of input frame data beyond the ability to be implemented by CPU processing or general GPU processing.

Embodiments herein describe a configurable engine integrated into a processor's cache hierarchy. The configurable engine can enable efficient data sharing between main memory, cache memory, and cores. The configurable engine can perform more efficient operations within the cache hierarchy. In one embodiment, the configurable engine is controlled (or configured) by software (e.g., an operating system (OS)) and is tailored to each application domain. That is, the OS can configure the engine according to the data flow profile of a particular application being executed by the processor.

Some embodiments of the teachings include: establishing a foundation layer comprising an OT domain low-code development tool, runtime, and a public information model, and an OT domain workflow generated by the development tool deployed on the runtime using the public information model; establishing a domain layer depending on the foundation layer, the domain layer comprising a field bus, a domain protocol, and a device tree of an application domain, the device tree comprising a connection relationship between devices; and establishing a user case layer that depends on the domain layer, the user case layer comprising user-defined function modules, and the user-defined function modules capable of being converted into the field bus, the domain protocol, and the device tree in the domain layer, wherein data obtained from a data platform is transmitted between the user case layer and the domain layer and transmitted between the layers using the model.

Systems and methods for evaluating a pre-trained artificial intelligence (AI) model using layered inputs. The system obtains a set of application domains in which the AI model will be used, and a set of guidelines that define one or more operational boundaries of the AI model. The system determines a set of layers, where each layer is associated with corresponding guidelines and mapped to a set of variables and benchmarks. Each variable represents an attribute within the guidelines and each benchmark indicates the degree of satisfaction of the AI model with the guidelines. The AI model is dynamically evaluated against these benchmarks using a series of assessments. Subsequent assessments are dynamically constructed based on the outcomes of previous assessments. Scores are assigned to the AI model for each layer by comparing the expected and actual responses. The results are then displayed in a graphical user interface (GUI).

This invention discloses a multimodal event joint extraction method based on a link matrix. It determines the event structure system of the application domain and constructs a cross-modal event extraction dataset. A link matrix between image patches and text sequences is constructed. A multimodal event extraction model is built based on ViLT, where the multimodal Transformer encoder takes image patch pixel features, word instance features, and image patch-text word instance pair features as input to obtain a multimodal feature input decoder, acquiring image-text event type consistency features, event type features, and image-text event element features. Using maximum likelihood estimation, the image-text event type matching result with the highest probability, the event type, and the image-text event element are taken as an index, and the text event information is obtained by index mapping. This invention solves the problems of low recognition accuracy in weakly supervised and unsupervised methods and inaccurate matching of pipeline-style multimodal event element information, providing a new event extraction strategy for the field of multimodal information extraction.

The invention discloses a method for quickly recovering a system after a host of a PCIE (Peripheral Component Interface Express) network fails, and aims at solving the problem of ensuring the timeliness and reliability of system recovery through a dual-host main-standby switching method when the host in the PCIE network fails. The method is established on a master-slave dual-computer communication structure taking PCIE (Peripheral Component Interconnect Express) exchange with a non-transparent bridge as a center, dual computers work normally at the same time and control own application domains, a slave computer monitors the online state of a master computer in real time through a heartbeat packet, and when the master computer fails, the slave computer automatically turns into the master computer state and takes over equipment resources managed by the master computer.

The present invention provides a question-answer retrieval method, apparatus, device and storage medium. The question-answer retrieval method, through the comprehensive application of word segmentation sequence, feature sequence and mixed sequence of question text, firstly, integrates application domain knowledge and language knowledge into the structural analysis of question text, so that the calculation results can be more in line with the expectations of business personnel. Secondly, by utilizing feature sequences, the training corpus in the corpus can be deduplicated, reducing the impact of changes in the specific values ​​of features on the calculation results. On the one hand, the accuracy of the calculation is improved, and on the other hand, the requirements for the amount of change in the corpus in the corpus are reduced, reducing the workload and difficulty of corpus maintenance. Thirdly, by calculating and adding word order weights, the word order factor is added on the basis of cosine similarity, which improves the accuracy of the calculation and sorting results and enhances the overall retrieval effect.

Provided in the present disclosure are an intelligent multi-domain efficient collaborative information transmission method and apparatus, and a storage medium. The method comprises: acquiring N data packets generated by an application domain; determining configuration information of the N data packets in a control domain on the basis of an artificial intelligence technique and a multi-domain joint optimization indicator, wherein the multi-domain joint optimization indicator includes at least one of the following: a throughput, a transmission delay, a weighted transmission delay and a quality of service; and transmitting the configuration information of the N data packets to at least one communication node, such that the at least one communication node determines a transmission resource on the basis of the configuration information of the N data packets, wherein the transmission resource is configured to transmit the N data packets.

This invention discloses a cost- and reliability-aware profit optimization method for a heterogeneous multi-server system. The method includes: establishing a heterogeneous multi-server system model to obtain expected service revenue and multi-server system utilization; establishing a soft error reliability model to formulate the application domain investment selection and profit maximization problem as a constrained nonlinear optimization problem; utilizing an iterative method based on a penalty function to obtain the optimal multi-server system configuration, minimum cost, and maximum profit for any single application domain, and converting the profit and configuration optimization problem of the heterogeneous multi-server system into a 0-1 knapsack problem; and utilizing a dynamic programming-based algorithm and a greedy-based algorithm to solve the optimal application domain investment strategy and the corresponding maximum profit. This method can consider budget constraints and the soft error reliability of service requests while configuring the heterogeneous multi-server system configuration, thereby improving cloud service quality and provider profits.

The invention relates to a data synchronization method and device, electronic equipment and a storage medium, and the method comprises the steps: receiving a detection synchronization message for a target application in a multi-application cooperation system, sending a fault detection instruction to the target application, enabling the target application to carry out fault detection, and obtaining fault detection data; the detection synchronization message comprises a message propagation flag bit, a cross-domain flag bit and a to-be-synchronized application domain identifier; based on the message propagation flag bit, the cross-domain flag bit, the target address and / or the to-be-synchronized application domain identifier, determining a target synchronization application corresponding to the fault detection data from a multi-application cooperative system comprising a target detection application domain and a to-be-synchronized application domain; the fault detection data is sent to the target synchronization application, so that the target synchronization application executes the association operation based on the fault detection data, the message transmission range can be precisely controlled, the granularity and flexibility of communication are improved, and the fault detection efficiency is improved.

A data processing device (1901) for end-to-end user query correction and slot filling is described, the device having one or more processors for: receiving (1801) a user query (301); processing (1802) the user query to determine whether the user query is correct for an application domain language or a natural language; forming (1803) a corrected query (350) based on the processing of the user query; performing (1804) a slot fill operation on the corrected user query to identify one or more slots (306) of the corrected query, where each identified slot corresponds to a portion of the corrected user query; each identified slot of the corrected query is classified (1805) as one of a plurality of categories. In addition to the slot filling task, this approach can handle incorrect input variants in multi-language settings and can handle application domain languages in addition to different natural languages.

The invention discloses a system software starting upgrading method and device and a vehicle. The method comprises the following steps: starting a first boot loader in a first application domain; responding to the upgrade mode identifier received by the first boot loader, and if the upgrade mode identifier is a specified character, starting a second boot loader in a second application domain; obtaining the network firmware and storing the network firmware in a second running storage space in response to the fact that the second boot loader analyzes the specified character; starting an upgrade state kernel, calling the upgrade state kernel to start a first format root file system, and obtaining a to-be-upgraded firmware mirror image package through the first format root file system; and performing firmware flashing on the first storage device and the second storage device based on the to-be-upgraded firmware mirror image package so as to start upgrading of operating system software configured in the first application domain and the second application domain. According to the method, the inter-core transmission process of the heteronuclear firmware can be omitted, and the flashing efficiency is improved.

The application discloses a program processing method of a domain controller. The method is applied to the domain controller, the domain controller comprises at least two M cores, at least one A core, a first storage medium and a second storage medium, the first storage medium comprises a boot program of the domain controller, the second storage medium comprises M core application programs, a configuration file of the domain controller and A core related programs, and the method comprises the following steps: the A core checks whether each program in the second storage medium can be read and written; if each program in the second storage medium can be read and written, the A core updates to-be-updated programs in the second storage medium based on the configuration file and a downloaded update file; and if at least one program cannot be read and written, a first M core in the at least two M cores executes the boot program of the domain controller, formats the damaged program through the boot program of the domain controller, and updates the to-be-updated programs based on the configuration file and the downloaded update file.

The present invention discloses a method and device for building a micro-front-end application, and relates to the field of computer technology. A specific implementation of the method includes: defining a micro-application tag, and creating a shadow document object model through the micro-application tag to generate a micro-application web native component; in response to the micro-front-end application construction operation, obtaining static resources under the sub-application domain name according to the domain name address of the sub-application; inserting the static resources into the shadow document object model to render the elements of the sub-application into the shadow document object model to introduce the sub-application into the main application. This implementation can solve the selector conflict problem and the problem of separation of the sub-application from the main application DOM structure, and there is no need to monitor route changes, and it can be rendered anywhere on the page, thereby reducing the configuration threshold, and realizing that a line of code can render the micro-front-end application, with high communication efficiency and flexible implementation.

A generator for converting an input vector from a latent space to one or more records x of measurement data that is realistic with respect to a given application domain. The generator includes: a trained neural network that is configured to map the input vector to a set of distribution parameters that characterize a random distribution of realistic measurement data, where this random distribution is configured such that given said set of distribution parameters and at least one source of randomness, samples of realistic measurement data may be obtained; and a sampling module including a random or pseudo-random number generator as a source of randomness and configured to sample the realistic measurement data from the random distribution.

The present disclosure relates to the field of network technologies and discloses an application access control method and system, and a device, a medium and a program product thereof. The present disclosure provides an application access control method. The method includes: generating, by a security management application client, an access request after detecting access to a target application, and sending the access request to a central domain name system; performing, by the central domain name system, domain name resolution on the access request to obtain a target application domain name of the target application, and sending the target application domain name to an application gateway; and determining, by the application gateway, a target access policy for the target application based on a matching result of the target application domain name in a first configuration file, and controlling, based on the target access policy, a terminal device.

The invention discloses a method and device for constructing an application field knowledge base based on regular expression, and the method comprises the steps: dividing and sampling a question set to obtain standard sentences, and constructing a standard sentence set, a matched sentence set and an unmatched sentence set; and subsequently, generating corresponding answers and regular expressions for the standard sentences in the standard sentence set by using the first large language model and the second large language model, and retrieving questions in the matched sentence set through the regular expressions corresponding to the standard sentences in the standard sentence set. Mapping questions retrieved by the standard sentences in the matched sentence set to obtain answers and regular expressions, combining questions not retrieved by the standard sentences in the matched sentence set with the unmatched sentence set to form a newly added sentence set, and generating corresponding regular expressions for the questions in the newly added sentence set by using a second large language model, and the first large language model is utilized to generate the corresponding answers for the questions in the unmatched sentence set, so that the workload can be reduced, the generation efficiency is improved, and the common problems of time consumption and error proneness in allocation of standard sentences to the questions are solved.

The present application discloses a method, device, chip, and vehicle for upgrading the hardware domain of a chip. The method is applied to the chip and includes: if the first security domain determines that the chip enters the first upgrade mode, releasing the control of the first storage area of ​​the first security domain to the first application domain, where the first upgrade resource is stored; generating and storing first identification information used to indicate that the first upgrade resource has been released; in the first upgrade mode, the second security domain releases the control of its second storage area to the first application domain, where the second upgrade resource is stored; if the first application domain obtains the first upgrade resource and the second upgrade resource based on inter-core communication, performing an upgrade operation on the inactive partition of the first application domain based on the first upgrade resource, the second upgrade resource, and the third upgrade resource stored by the first application domain itself. This method implements the upgrade of a portion of the hardware domain.

The invention relates to the field of data processing, and particularly discloses a category activation method and device, electronic equipment and a storage medium. In response to a category activation request submitted by a user, a user domain module is called to obtain user information; in response to a first selection instruction for selecting at least one candidate account, calling an account domain module to obtain a category list corresponding to the selected target account; in response to a second selection instruction for at least one target category in the non-activated category list, calling the application field submitting module to perform preposition check on the at least one target category corresponding to the second selection instruction, and performing preposition operation when the at least one target category needs preposition operation; in response to submission information fed back by the application domain submission module, calling the user application domain module to activate and check at least one target category corresponding to the second selection instruction; and when the target category is activated and audited successfully, the activation result of the target category is fed back, so that the maintenance cost of the category activation function is reduced.

The invention provides a method for realizing visual programming, and belongs to the technical field of visual programming, and the method comprises the following steps: determining different field requirements according to application fields, listing a programming normal form and a programming block type based on the field requirements, and designing an overall architecture corresponding to the application fields; specific functions of the programming blocks are determined based on the overall framework, the connection relation between the programming blocks is determined according to the specific functions, and then a core module of the corresponding application field is developed; a user interface and user interaction mode is designed in combination with the overall framework and the core module, so that a visual programming environment is formed, the programming threshold is lowered, the programming process is simplified, the user experience is improved, non-professional users can easily master the method through visual programming, the development efficiency is improved, and the requirement for user environment configuration is lowered.