701 results about "Language model" patented technology

Methods, systems, and products for alert generation and augmentation using a large language model (‘LLM’), including: providing, to the LLM, information describing a monitored deployment; providing, to the LLM, information describing one or more alerts associated with the monitored deployment; and receiving, from the LLM, an enhanced alert, wherein the enhanced alert is based at least in part on at least one of the information describing the monitored deployment and the information describing one or more alerts associated with the monitored deployment.

Systems and methods are disclosed herein for receiving, based on user interaction with a user interface, a user input of a natural language search query for identifying a cybersecurity threat by way of a search interface, the natural language search query requesting a specialized search of a threat database. An application generates a search vocabulary based on the natural language search query. The application performs a query lookup using the threat database, the query lookup returning a plurality of files that at least partially match the search query. The application prompts a large language model to generate an answer to the natural language search query using the plurality of files that at least partially match the search query, and outputs for display the answer using the user interface.

Techniques for automatically de-identifying sensitive information in audio conversations by combining un-transcribed voice activity detection (VAD) with large language model (LLM) analysis are disclosed. An audio de-identification system processes speech-to-text transcriptions while identifying segments where automatic speech recognition (ASR) failed to transcribe spoken content. These un-transcribed segments are represented as placeholders in prompts sent to an LLM, which analyzes the surrounding textual context to determine if sensitive information (such as PII or PHI) was likely spoken during these gaps. When sensitive content is identified, the system modifies the corresponding audio segments through an audio identification tactic. This approach addresses the technical challenge of incomplete de-identification in automated audio processing by leveraging LLMs' contextual understanding to detect sensitive information in segments that traditional ASR systems miss, particularly in scenarios involving poor audio quality or diverse accents. The result is a more comprehensive and reliable audio de-identification system.

One embodiment sets forth a technique for fine-tuning a machine learning model to perform physical reasoning. According to some embodiments, the method can include the steps of obtaining simulation annotations that describe interactions among simulated objects within a physics-based environment and one or more question templates, each question template defining a different parameterized reasoning query; generating, based on the simulation annotations and the one or more question templates, a plurality of question-answer pairs that represent physical reasoning examples; formatting the question-answer pairs into natural-language data compatible with the machine learning model; and fine-tuning the machine learning model based on the natural-language data.

A text sequence recommendation method and system based on a large language model is disclosed, belonging to the technical field of recommendation algorithms. The method includes: a data preprocessing stage, a large language model pre-training stage, a sequence model fine-tuning stage and a matching stage. According to this disclosure, a large language model is introduced into a text sequence recommendation task, so that text can be better modeled by utilizing rich pre-training corpus of the large language model; meanwhile, sequence modeling is performed on the text, the capability of sequence recommendations modeling in a large model is activated, an ID-based recommendation paradigm in a traditional recommendation algorithm is eliminated, and recommendation task learning processing is better performed in a cold start scenario and a knowledge transfer scenario; and finally, a recommendation result is finally optimized by a sequence model.

Anomalies regarding a specified entity that occurred within a specified time period are selected from anomalies identified by security analysis performed on raw events. The selected anomalies are enhanced with additional information regarding the selected anomalies. A prompt is generated based on the selected anomalies as have been enhanced. The prompt is generated to solicit a response from a large language model (LLM) including a natural language summary synthesizing the selected anomalies. The generated prompt as input to the LLM, and the response is received as output from the LLM.

A first large language model (LLM) instance may be instructed to request data while being prevented from performing calculations using the data. A second LLM instance may be instructed to provide a response to the request for data based on a known complete data set. The response may be translated into a machine-readable response in a format configured for processing by a calculation engine. The calculation engine may process the machine-readable response, thereby generating a calculation engine output. A mismatch between the calculation engine output and a known result obtained using the known complete data set may be identified, and the instruction to the first LLM may be modified in response.

A computer-implemented method includes: receiving, by a computing device, an input prompt; routing, by the computing device, the input prompt to one or more restructuring routes; restructuring, by the computing device, the input prompt to provide a restructured input prompt to reduce tokens; evaluating, by the computing device, the input prompt to the restructured input prompt to determine that thresholds are met; and sending, by the computing device, the restructured input prompt that meets the thresholds to a large language model.

A signal generation system includes a first computing device configured to generate a trajectory signal representing a planned trajectory of a vehicle; and a second computing device configured to generate an output word signal representing word information to be given to an occupant of the vehicle, by inputting an input word signal representing an utterance of the occupant in natural language format into a trained language model. At least one of the first computing device and the second computing device converts at least one of the sensor signal and a signal generated based on the sensor signal to a trajectory data signal represented in a format that can be inputted into the trained language model. The second computing device generates a trajectory word signal related to the planned trajectory as the output word signal, based on the trajectory data signal.

The application discloses a kind of intelligent query control method and electronic equipment based on natural language interaction.The method includes the following steps: building context knowledge base, converting into context vector, and storing in vector database;Convert the input natural language question into query vector;Retrieve the k context vectors most similar to the query vector from the vector database, and determine the corresponding context information;Combine the natural language question and the retrieved context information into a prompt word sequence input into a large language model (LLM) to obtain an SQL query statement, and analyze whether the natural language question contains a control intent;Execute the SQL query statement;If the natural language question contains a control intent, analyze the query result intelligently and perform the corresponding device control operation.The application can analyze the control intent in the natural language question, and after the result is queried, the corresponding device can be controlled according to the control intent to work, improving the work efficiency.

A control circuit receives a user query and inputs that query to at least one model-based query classifier and outputs a selection one or more information resources to provide a selected information resource. The control circuit can be further configured to assess the aforementioned user query to determine context sufficiency (for example, by employing predetermined large language model prompts to determine the context sufficiency). When the context sufficiency is determined to be sufficient, the control circuit can provide the user query as an output query and direct the output query to the selected information resource. When, however, the context sufficiency is determined to be insufficient, the control circuit can rephrase the user query to include additional chat history to serve as the output query and then direct that output query to a selected information resource corresponding to a nearest prior query on a same query sequence.

A computer-implemented system and method relates to language-guided self-supervised semantic segmentation. A modified image is generated by performing data augmentation on a source image. A machine learning model generates first pixel embeddings based on the modified image. First segment embeddings are generated using the first pixel embeddings. A pretrained vision-language model generates second pixel embeddings based on the source image. Second segment embeddings are generated by applying segment contour data from the first pixel embeddings to the second pixel embeddings after the data augmentation is performed on the second pixel embeddings. Embedding consistent loss data is generated by comparing the first segment embeddings in relation to the second segment embeddings. Combined loss data is generated that includes the embedding consistent loss data. Parameters of the machine learning model are updated based on the combined loss data.

The application provides a semantic extraction method based on model computing power dynamic adjustment, a model and computer equipment. The method comprises the following steps: calling a plurality of feature extraction layers to perform multiple feature extractions on initial semantic features to obtain predicted semantic features; judging the matching degree of the predicted semantic features and preset expected semantic features; if the matching degree of the predicted semantic features and the preset expected semantic features meets a preset matching threshold, generating an output result of a language model through the predicted semantic features; and if the matching degree of the predicted semantic features and the preset expected semantic features does not meet the preset matching threshold, taking the predicted semantic features as new initial semantic features, re-calling the plurality of feature extraction layers to perform multiple feature extractions on the new initial semantic features until the output result of the language model is generated. The embodiments of the application can allocate different computing resources for different tasks, so that the purpose of reasonably using computing resources is achieved.

The present application relates to the technical field of remote sensing image processing and artificial intelligence, in particular to a state space model driven remote sensing image change caption generation method, the method comprises the following steps: obtaining image data and word mapping table of remote sensing image intelligent interpretation; extracting double time phase features through a double time sharing branch feature extractor; constructing a joint input sequence of a decoder according to the difference enhanced features, multi-scale visual features and text sequences; taking the constructed joint input sequence as input, calculating word probability distribution by autoregressive calculation based on a pre-trained language model decoder, completing prediction and cyclic iteration according to the word mapping table, and obtaining complete image change description sentences; the present application solves the problem of losing high frequency details in long sequence modeling of state space model through frequency domain modulation, and effectively suppresses background noise interference by using difference perception aggregation, thereby improving the accuracy of remote sensing image difference description generation.

The application provides a kind of operation and maintenance scene visualization report generation method and system based on natural language question and answer, belongs to intelligent operation and maintenance technical field, method includes: receiving the natural language query input by user;Through the pre-training of large language model and the preset operation and maintenance terminology dictionary, the natural language query is parsed and entity is extracted, and the entity-field association table containing demand type is generated;According to demand type and semantic feature, the query type is judged to be data query or visual query;Based on the judgment result, generate structured query language sentence or visual instruction;Query is executed to the interface business database to obtain raw data;Raw data is processed and analyzed to obtain analysis result data;Call data feature adaptation algorithm to match chart type, generate and output visual report.The application realizes the full-link automation from natural language input to visual report output, reduces the operation and maintenance data interaction threshold, improves the operation and maintenance data processing efficiency and accuracy.

This device for generating crystal structure information on the basis of an artificial intelligence model comprises: a data preprocessing unit that constructs an interleaved dataset by arranging text and images to cross each other; a learning unit that includes a visual encoder for generating image feature data from the images, a language model for generating text-encoded data from the text, and a projector for converting the image feature data and the text-encoded data into data in a shareable form, learns a relationship between the images and the text on the basis of the converted data, initializes the projector, and then performs, on the basis of the interleaved dataset, fine tuning for simultaneously training the language model and the projector to perform condition learning for generating information, which satisfies a specific condition, and filling learning in a method for filling a masked portion; and an information generation unit for generating at least one piece of crystal structure information on the basis of a prompt input.

The application discloses a multi-round multi-modal toxicity detection method and device, and relates to the technical field of natural language processing and computer vision. The method comprises the following steps: obtaining multi-round dialogue data of a Chinese social platform, constructing a multi-modal dialogue data set containing text and images and performing preprocessing to obtain a processed multi-modal data set; designing a multi-dimensional labeling system, performing multi-dimensional toxicity type labeling on each dialogue, and calculating a labeled real cumulative toxicity score; constructing a systematic evaluation framework, and setting a toxicity content evaluation task; based on a test sample data set and the set toxicity content evaluation task, performing toxicity detection on the test sample data set by using a pre-trained visual language model, outputting a cumulative toxicity score predicted by the pre-trained visual language model, and calculating a cumulative toxicity detection accuracy; and based on the cumulative toxicity detection accuracy, performing evaluation, and outputting a model evaluation result. The application can improve the evaluation of original visual content understanding ability.

The present invention provides an information processing device, method, and program that perform highly accurate item extraction while reducing computational complexity, even when item extraction is performed on documents where the evaluation values ​​of multiple candidates for item names or item values ​​are approximately the same, or on documents that satisfy given conditions. [Solution] The method involves performing character recognition processing on a document image to obtain character information about the strings contained in the document image, extracting candidate strings corresponding to at least one of the item names and item values ​​to be extracted from the character information, obtaining an evaluation value indicating the likelihood of the candidates, and if the evaluation value of the candidates satisfies predetermined conditions, extracting the strings corresponding to at least one of the item names and item values ​​contained in the document image using a large-scale conversational language model (LLM).

This application discloses an open-world 3D target detection method and apparatus, comprising: acquiring multimodal data of a target scene; performing 3D target detection on 3D point cloud data to obtain 3D candidate targets; inputting the geometric features of each 3D candidate target into an out-of-distribution target classifier to determine whether the 3D candidate target belongs to a known category set and to identify unknown targets; projecting the 3D candidate targets corresponding to the unknown targets onto 2D image data, obtaining the 2D image region, and inputting it into a visual language model, using natural language prompts to guide the visual language model to output the semantic category name of the unknown targets; fusing the semantic category name and the spatial location information of the 3D candidate targets to generate open-world 3D detection results. This application can effectively identify unknown 3D targets in an open world and generate open-world 3D detection results that combine spatial positioning and semantic description.

In some embodiments, a secure chatbot includes a generative language module (GLM) and a prompt manager configured to dynamically update GLM input prompts according to outputs of the GLM. The input prompts instruct the GLM to perform a task (e.g., determine whether a target message received by a user indicates online fraud) and, in response, output a particular token. In response to detecting the token within the output of the GLM, the prompt manager can selectively execute a code snippet identified according to the respective token. Executing the code snippet effectively updates the prompt, which is then fed back to the GLM for a next inference step. In contrast to conventional LLM prompts, the updates to the prompts herein include modifications (e.g., insertion of other tokens, etc.) rather than merely appending inferred tokens to previous prompts. Moreover, the prompt modifications are executed inline, i.e., without submitting a wholly new prompt to the GLM.

A computer-implemented method of generating verification data for a query result provided by a large language model, LLM, includes generating a prompt for the large language model. The prompt contains a verification request for a query, the query including query text and input data from which the query result can be derived. The verification request includes instructions that cause the LLM to generate verification data that indicates a derivation of the query result from the input data. Another computer-implemented method includes receiving the verification data and processing the verification data to determine whether the query result was validly derived from the input data.

The application provides a pre-training language model construction method, system and device, relates to the fields of artificial intelligence technology and semantic processing, and mainly comprises the following steps: based on a knowledge graph, a subgraph with a triple as a node unit is constructed; based on entity and relation description text, both are encoded through a conventional pre-training language model to obtain entity representation vectors and relation representation vectors; for the entity representation vectors and the relation representation vectors corresponding to the subgraph, modeling relation information is carried out based on a graph self-attention neural network, and the parameters of the graph neural network and the entity vectors are updated through a prediction link task; the parameters of the model are updated based on a contrast learning technology; and the parameters of the model are fine-tuned for a specific task. The scheme overcomes the problem that entity in the equipment field is sparse and a conventional pre-training language model cannot sufficiently learn entity semantics, and the updated pre-training language model has better understanding and reasoning capabilities in the application of the equipment field.

The application discloses a demand-driven visual language model rule generation method and device, which comprises the following steps: receiving a natural language demand of a user, analyzing the natural language demand into a structured constraint template, converting the structured constraint template into an executable rule and binding a rollback point; executing the executable rule on a current sample and a same batch sample based on a visual language model, outputting a tabular report and a corresponding evidence package; rejudging suspicious samples in the tabular report, judging whether false positives or false negatives occur, updating the rule according to a rejudging result, and recording a version change log. The scheme significantly improves the accuracy and flexibility of rule generation.

A method for training a student model based on a pre-trained language model (PLM) and a text classification system are disclosed. The method includes: constructing cue-based training samples; adjusting the pre-trained language model using the cue-based training samples to obtain a cue-adjusted teacher model; and training the student model using the processed training samples, wherein during training, the student model simultaneously learns the classification probability vectors output by the cue-adjusted teacher model and the original teacher model. This invention requires the student model to learn from two teacher models simultaneously, thereby alleviating the overfitting problem of the student model in small-sample scenarios by adding a distillation path that learns from the original PLM teacher model with unsupervised data. Furthermore, by transferring the intermediate layer representation of the PLM through knowledge probes and stabilizing the performance of knowledge distillation through comparative learning, the student model can learn higher-order dependencies from the intermediate layer representation of the teacher model, improving the accuracy and efficiency of knowledge distillation.

This application discloses an intelligent retrieval control method, device, and electronic device based on large model dimensionality reduction, relating to the field of artificial intelligence technology. The method includes: inputting the text to be retrieved into a large language model as an embedding model, and obtaining the output vector of a specified hidden layer of the large language model to obtain a text vector to be retrieved, the dimension of which is M; using a specified window to perform dimensionality reduction processing on the text vector to be retrieved, reducing the text vector to N-dimensionality reduced text vectors, the window size of which is determined based on M and N; dividing the text vector to be retrieved into N sub-vectors by the specified window; in the text vector to be retrieved, the corresponding element positions of adjacent sub-vectors in the N sub-vectors are at least partially different; each element in the dimensionality reduced text vector is the average of the elements in one of the N sub-vectors; and performing vector retrieval using the dimensionality reduced text vector.

The application provides a PCB quotation parameter extraction method, system, device and medium, comprising: converting the obtained source file into a picture; dividing the picture into corresponding file types; classifying a user-defined to-be-extracted parameter list to obtain a file source-parameter group-knowledge prompt word dictionary; extracting filter information prompts from the filtered files of the source file; inputting the file source-parameter group-knowledge prompt word dictionary and the filter information prompts into a visual-language model to output a final PCB quotation parameter set.

To appropriately assess customer emotions in a call center. [Solution] An information processing device comprising: an acquisition unit that acquires text data indicating the content of a conversation between a customer and an operator in a call center; an evaluation unit that inputs evaluation data based on the text data and a prompt including a request to select from multiple candidates that categorize the customer's emotions into a large language model (LLM) and acquires the output of the large language model to evaluate the customer's emotions; and an output unit that outputs the processing result of the evaluation unit.