21 results about "Context manager" patented technology

The invention provides an agent collaborative question-answering method and system based on context awareness and authority control, and belongs to the technical field of artificial intelligence. The method comprises the following steps: acquiring user query, and triggering a user context manager to generate a session state user context object containing a permission label and preference; the central scheduler performs task decomposition based on the query intention and the context, dynamically selects an intelligent agent through a multi-dimensional scoring model and binds a knowledge source; the authority-aware agent execution engine executes triple authority verification and intermediate result desensitization processing in the whole process, and supports agent fault tolerance and audit log recording. The method realizes unification of knowledge security and access flexibility, supports multi-scene adaptive question and answer, and has good expandability and fault-tolerant capability.

This disclosure describes a large language model (LLM)-based agent operating system comprising an application layer, a kernel layer, and a hardware layer. The kernel layer comprises an AIOS kernel comprising: LLM core(s), an agent scheduler, a context manager, a memory manager, a storage manager, and a system call interface configured to manage interactions among the agent scheduler, the context manager, the memory manager, and the storage manager.

A direct memory access controller (DMAC) includes a virtual channel, a physical channel, and a context manager. The virtual channel is configured to generate a flow control signal for executing a direct memory access (DMA) command. The physical channel includes read and write control logic to initiate data transmission from the source device to the destination device in accordance with the DMA command in response to the flow control signal. In one embodiment, a context manager includes: allocation logic configured to arbitrate between allocation requests from virtual channels and to allocate physical channels to selected virtual channels; and routing logic configured to route a flow control signal from the selected virtual channel to the allocated physical channel, and to route a status signal between the allocated physical channel and the selected virtual channel.

The application discloses a kind of microservice dynamic access control strategy under zero trust architecture method and device, belong to network security, distributed system, zero trust, microservice field, it specifically includes: by the event tracker of node to obtain the context information in microservice runtime context event, context manager collects and aggregates context information and is reported to microservice control face global manager, global manager is according to global context mapping table value, judge whether it is new microservice instance, if yes then update global and local Tag mapping table, when node receives message, message verifier obtains the Tag value carried by this message, by local Tag mapping table to obtain the microservice context information corresponding to this message, after inquiring local strategy table and global strategy table, run decision algorithm, execute corresponding access control strategy, use context to identify communication endpoint, with dynamic, accurate, scalable advantage.

The invention relates to an electronic medical record quality evaluation and verification method, which comprises the following steps of: S1, cleaning, analyzing and standardizing a complete electronic medical record to obtain a standardized patient panoramic context and a text block set capable of being retrieved by a vector; s2, constructing a mixed quality control knowledge base according to the external medical knowledge base and the quality control rule set; and S3, inputting the patient panoramic context, the mixed quality control knowledge base and a to-be-executed quality control point into a dynamic context manager, checking the document sequence set by adopting a quality evaluation function, and outputting a structured defect report. According to the method, an electronic medical record quality evaluation and verification framework integrating dynamic context management, multi-source knowledge fusion and large-scale language model reasoning ability is constructed, and the core contribution of the method is that the context window limitation of LLM is effectively avoided by designing an exquisite context manager mathcal {M}, and the reliability of the system is improved. And therefore, deep and explainable logic review can be executed on high-dimensional and time-sequence EMR (Empirical Mode Register) data.

The application discloses a kind of text context dynamic reconstruction methods, comprising: to original document is step length overlap segmentation processing, generate semantic fragment and record physical offset index information;Build document context manager storage original document complete content;Perform vector retrieval and obtain the physical offset index of matching fragment;Real-time analysis and determine the physical position boundary of fragment in original document;From original document, real-time extraction predecessor and successor text block;Splice form complete logical context window.The application establishes the accurate position anchor point of semantic fragment and original document by physical offset index, realizes zero-cost backtracking original;Adopt retrieval time dynamic reconstruction, support according to the context expansion range of adaptive adjustment of retrieval confidence;Provide three-level context mode to meet different retrieval scene demand;Intelligent deduplication eliminates redundant content.Compared with traditional RAG system, retrieval accuracy is improved by about 18%, context completeness is improved by about 31%, and Token consumption is reduced by about 30%.

The invention discloses a multi-tenant container cloud-oriented RDMA performance isolation system. A monitoring module comprises an eBPF monitoring program and a global context manager; the control path scheduler is deployed in a user mode library of a tenant and is used for receiving feedback information of the monitoring module and limiting or rejecting the original intonation rate of the control path exceeding a threshold value; and the data path scheduler is used for shaping and scheduling a data path task request issued by the tenant application program so as to control the sequence and the speed of entering the RDMA network card. According to the method, the eBPF monitoring program is adopted to monitor the original intonation frequency of the control paths of all tenants and the resource use condition of the RDMA network card, and millisecond-level response to RDMA control path operation is achieved. Meanwhile, a dynamic task segmentation mechanism based on active delay is adopted, and RDMA network card resource occupation of different tenants is balanced; and fairness is ensured based on a scheduling strategy of admission control.

A processing core for executing threads is provided. The processing core comprises: one or more functional units; a shared register file comprising a plurality of registers, each register of the shared register file being configured to store data for executing any thread; and a multi-thread elaboration unit, MTEU, the MTEU comprising: a control module configured to: receive an indication of a plurality of active threads for a given scheduling window; and receive instructions for each of the plurality of active threads; a context manager module configured to load, for each of the plurality of active threads, data required for executing a next one or more of the instructions into a respective one or more registers of the shared register file; and an instruction scheduler module configured to schedule the instructions of the active threads for execution based on availability of the one or more functional units.

The application discloses a kind of text context dynamic reconstruction methods, comprising: to original document is step length overlap segmentation processing, generate semantic fragment and record physical offset index information;Build document context manager storage original document complete content;Perform vector retrieval and obtain the physical offset index of matching fragment;Real-time analysis and determine the physical position boundary of fragment in original document;From original document, real-time extraction predecessor and successor text block;Splice form complete logical context window.The application establishes the accurate position anchor point of semantic fragment and original document by physical offset index, realizes zero-cost backtracking original;Adopt retrieval time dynamic reconstruction, support according to the context expansion range of adaptive adjustment of retrieval confidence;Provide three-level context mode to meet different retrieval scene demand;Intelligent deduplication eliminates redundant content.Compared with traditional RAG system, retrieval accuracy is improved by about 18%, context completeness is improved by about 31%, and Token consumption is reduced by about 30%.

The invention discloses a transaction control system and method for a database. As an openGauss kernel extension module, the system is composed of a session transaction context manager, an SQL statement classification interceptor, a DML transaction state machine engine, a DDL transaction coordinator and a PL / pgSQL transaction boundary analyzer. The system supports session-level automatic submission mode configuration, a DML transaction block is automatically started through a state machine in a non-automatic submission mode, and multi-statement atomicity can be achieved without depending on BEGIN statements; the DDL operation is independently and automatically submitted according to a strategy or is decoupled from a DML transaction for execution; and the PL / pgSQL storage process analyzes a forced explicit transaction boundary through a static code. According to the method, transaction management codes are simplified, atomic damage caused by omission of BEGIN is avoided, unified arrangement of DDL and DML is supported, the robustness of the storage process is improved, and the method is transparent and compatible with standard application and has good expandability.

The invention discloses a multi-source heterogeneous JSON (JavaScript Object Notation) data stream processing method based on Flink SQL (Structured Query Language), which specifically comprises the following steps of: establishing a Kafka connection table, defining a Topic field and an original data field extracted from Kafka metadata in a table structure of the connection table, and starting a dynamic Schema mode; creating an independent structure context manager for each physical processing task instance of the Flink SQL, wherein the manager maintains structure contexts of different Topic in a key value pair form; analyzing the key value pair of each JSON, performing key name standardization processing, and executing newly added field addition or type conflict processing based on the Schema cache; recursively unfolding the object and the array type field, and generating an unfolding structure according to the maximum unfolding depth limitation and the position marking rule; constructing an output line according to the Schema cache, filling a missing field into a null value, and injecting a Kafka metadata field and an expansion position mark; and outputting Row objects in a uniform format to form a dynamic Schema data stream, and butting an Flink Table API (Application Program Interface) to support direct query of a dynamic extension field.

The invention relates to an alarm event processing method and device in a server, computer equipment and a medium, and the method comprises the steps: responding to an alarm system configuration operation, and transmitting configuration dynamic data corresponding to the alarm system configuration operation to a jLogtail collector; acquiring an original log of a server collected by a jLogtail collector, converting the collected original log into a standard format event, and pushing the standard format event to a message queue; calling an MCP context manager to consume the standard format event from the message queue, supplementing context information by taking a core identifier corresponding to the standard format event as a clue, and generating complete context information; and based on the complete context information, in combination with a dynamic threshold algorithm and a weighted diagnosis result of the diagnosis agent cluster, judging whether a preset alarm triggering condition is met. The whole scheme can realize efficient and accurate alarm event processing.

This invention discloses a method and system for small-sample intent assessment and situational awareness based on a large language model, belonging to the field of natural language processing. It mainly includes: retrieving the most relevant few-sample examples and known intent descriptions for queries using a dynamic context manager; concatenating prompts and invoking a large language model for inference, outputting intent recognition results and discovering new intents; employing a joint training strategy of supervised contrastive learning and adversarial enhancement to refine the feature space on pseudo-labeled data; and calculating anomaly scores for samples based on Mahalanobis distance to identify out-of-distribution new intents (abnormal intents). This invention, through a two-stage hybrid framework of "generative guidance and discriminative refinement," effectively solves the technical challenges caused by data scarcity and dynamic evolution of intent in open environments, achieving highly robust, low-cost, and sustainably evolving intent discovery and recognition. It can be widely applied in small-sample fields such as military activities, cybersecurity, financial risk control, and public opinion analysis.

The following descriptions concern a system for providing and interacting with a context-aware omnichannel vehicle assistant, comprising: a generic context manager implemented on a server for receiving and processing data from a variety of entities to collect real-time data on the vehicle status, the driver's condition, environmental conditions, and / or driver interactions across various touchpoints; a first local context manager implemented in a first entity and a second local context manager implemented in at least one further entity for local data processing; and at least one output unit.wherein the generic context manager includes a first machine learning and AI module for using machine learning algorithms to analyze the driver's preferences, predict the driver's needs, and adapt services accordingly, wherein the first learning and AI module enables a personalized driver experience and predictive adaptations, wherein the generic context manager communicates with the first local context manager and the second local context manager via an omnichannel interface for exchanging data and / or integrating the first entity with the at least one other entity.

The application provides a code block-based model precision alignment test method and system and electronic equipment, and relates to the technical field of deep learning testing. The method comprises the following steps: defining an execution boundary for a code block to be tested by a context manager and monitoring; based on input data in a reference calculation record prepared for the code block, executing the code block under monitoring, and automatically capturing a generated forward data set; the forward data set at least includes an input tensor input to the code block and an intermediate tensor generated inside the code block; after the completion of reverse gradient calculation, a gradient data set corresponding to the forward data set is automatically captured; finally, the captured forward data set and / or gradient data set is compared with corresponding data in the reference calculation record, and a precision alignment test report is generated. The application realizes fine-grained and low-invasive precision testing of the calculation process of any code segment in a deep learning model, and solves the problem of coarse testing granularity and poor flexibility of traditional methods.

The application provides a type inference method and device of a programming language, the method comprising: in the development process of program code, obtaining a node to be inferred in an abstract syntax tree, the node to be inferred indicating a field to be inferred in the program code, the field to be inferred comprising one of the following: a variable, a class, a function or a parameter; obtaining a constraint related to the node to be inferred, the constraint being used to indicate a relationship involved by the node to be inferred in the abstract syntax tree; determining a type of the node to be inferred according to the constraint, the type of the node to be inferred being one of the following: a numeric type, a sequence type, a mapping type, a set type, an iterator type, a context manager type, a module type, a metatype, a null type or a callable type. The scheme of the application is beneficial to reducing the memory occupation of type inference.

The invention discloses a tile type region cutting path correction method and system based on context prediction, and relates to the technical field of path correction. After reference mark points and scanning mark points recorded in a preset context manager are preprocessed through space coordinates, missing mark point prediction compensation is carried out based on a parallelogram geometric rule, and a target quadrangle and a source quadrangle are constructed; associating the target quadrangle and the source quadrangle of each spatial position, generating a tile unit, and determining a local transformation matrix corresponding to each tile unit; and correcting the coordinates of the key points on the cutting path according to the local transformation matrix corresponding to each tile unit to obtain the coordinates of the target points, and determining the cutting path according to all the corrected coordinates of the target points. The problem that high-precision cutting path correction can still be achieved under the conditions that large-breadth non-uniform deformation and mark point adding are limited is solved.

A direct access memory controller (DMAC) includes virtual channels (VCHs), physical channels (PCHs) and a context manager. A VCH is configured to generate a flow control signal for execution of a dire

The invention discloses an agent SQL generation system based on reinforcement learning enhancement, and belongs to the technical field of artificial intelligence. The intelligent database query generation system is light in weight, safe and capable of being deployed privately. A decision-making layer is respectively connected with a unified tool layer and a context management layer; a user inputs questions to the business knowledge vector knowledge base and then splices the questions to a new question obtained by the large language model as input of a decision-making layer; in the decision-making layer, a core execution mechanism is used as core decision-making logic, and decision-making tasks are executed according to the staged workflow; a context manager with stage perception is arranged in the context management layer and is based on a stage context compression mechanism; the unified tool layer is responsible for interacting with all databases through interfaces of the unified tool layer; the decision-making layer generates a final database query statement after completing the decision-making task, inputs the final database query statement into a database for operation through the unified tool layer, obtains a corresponding result, carries out decision-making submission, and carries out chart visualization processing to obtain a query result.

A direct memory access controller (DMAC) includes virtual channels, physical channels and a context manager. A virtual channel is configured to generate a flow control signal for execution of a direct memory access (DMA) command. A physical channel includes read and write control logic to initiate, responsive to the flow control signal, transfer of data from source device to a destination device in accordance with the DMA command. The context manager includes allocation logic configured to arbitrate between allocation requests from the virtual channels and allocate a physical channel to a selected virtual channel and routing logic configured to route a flow control signal from the selected virtual channel to the allocated physical channel and to route a status signal between the allocated physical channel and the selected virtual channel.