24 results about "Representation language" patented technology

Approaches presented herein provide for the use of language models to generate tokenized descriptions of physical environments. In at least one embodiment, sensor and / or observational data can be obtained for an environment and used to generate a set of perception data. The perception data can be analyzed, along with approximate positional data within the environment, to identify a set of aligned map data. The aligned map data and perception data can be provided as input to a trained language model, which can be trained to correlate and / or fuse the information to generate a single, consistent representation of the environment. The language model can output a tokenized description of the environment, which can be in a domain-specific language, that is a compact but robust textual description of the environment.

Techniques for augmenting training data include accessing training data comprising a plurality of training examples comprising a first training example comprising a first natural language utterance and a first logical form for the first natural language utterance. A second natural language utterance is generated by adding or replacing one or more values in the first natural language utterance. A logical form for the second natural language utterance is generated. A second training example is generated, comprising the second natural language utterance and the logical form for the second natural language utterance. The training data is augmented by adding the second training example to the plurality of training examples to generate an augmented training data set. A machine learning model is trained to generate logical forms for utterances using the augmented training data set.

Techniques are disclosed herein for improving model robustness on operators and triggering keywords in natural language to a meaning representation language system. The techniques include augmenting an original set of training data for a target robustness bucket by leveraging a combination of two training data generation techniques: (1) modification of existing training examples and (2) synthetic template-based example generation. The resulting set of augmented data examples from the two training data generation techniques are appended to the original set of training data to generate an augmented training data set and the augmented training data set is used to train a machine learning model to generate logical forms for utterances.

The present disclosure relates to manufacturing training data by leveraging an automated pipeline that manufactures visualization training datasets to train a machine learning model to convert a natural language utterance into meaning representation language logical form that includes one or more visualization actions. Aspects are directed towards accessing an original training dataset, a visualization query dataset, an incremental visualization dataset, a manipulation visualization dataset, or any combination thereof. One or more visualization training datasets are generated by: (i) modifying examples in the original training dataset, the visualization query dataset, or both to include visualization actions, (ii) generating examples, using the incremental visualization dataset, the manipulation visualization dataset, or both, that include visualization actions, or (iii) both (i) and (ii). An augmented training dataset is generated by adding the one or more visualization training datasets to the original training dataset and then used to train the machine learning model.

The invention discloses a multi-source heterogeneous model fusion method and device, a storage medium and electronic equipment, and belongs to the technical field of computer intelligence. The method comprises the following steps: acquiring a multivariate heterogeneous sub-model to be fused; analyzing the sub-models to obtain personalized features of the multivariate heterogeneous sub-models to be fused, and constructing a unified semantic representation system based on a meta-model framework; according to the personalized features of the sub-models, standardized expression is carried out on the logic expressions of the sub-models by utilizing a unified intermediate expression language to obtain a to-be-fused multivariate heterogeneous sub-model with a standardized expression form; and according to the sub-model with the standardized expression form, a multivariate heterogeneous model based on the to-be-fused multivariate heterogeneous sub-model is constructed and obtained based on the knowledge graph and the graph neural network. The device, the storage medium and the electronic equipment can be used for implementing the method, and the efficiency and reliability of multi-model collaborative design and simulation verification in the development process of complex equipment products can be remarkably improved.

The present disclosure is related to techniques for converting a natural language utterance to a logical form query and deriving a natural language interpretation of the logical form query. The techniques include accessing a Meaning Resource Language (MRL) query and converting the MRL query into a MRL structure including logical form statements. The converting includes extracting operations and associated attributes from the MRL query and generating the logical form statements from the operations and associated attributes. The techniques further include translating each of the logical form statements into a natural language expression based on a grammar data structure that includes a set of rules for translating logical form statements into corresponding natural language expressions, combining the natural language expressions into a single natural language expression, and providing the single natural language expression as an interpretation of the natural language utterance.

Techniques are disclosed herein for improving model robustness on operators and triggering keywords in natural language to a meaning representation language system. The techniques include augmenting an original set of training data for a target robustness bucket by leveraging a combination of two training data generation techniques: (1) modification of existing training examples and (2) synthetic template-based example generation. The resulting set of augmented data examples from the two training data generation techniques are appended to the original set of training data to generate an augmented training data set and the augmented training data set is used to train a machine learning model to generate logical forms for utterances.

Techniques are disclosed herein for calibrating a confidence score of a machine learning model trained to convert a natural language into a meaning representation language. The techniques include obtaining one or more original bundling scores generated from one or more bundling levels of a decoder trained as a machine learning model that translates natural language to logical form, each of the one or more original bundling scores is a conditional probability of a sub-tree determined by a heuristic search algorithm of the decoder at one of the one or more bundling levels; classifying, by the calibration model, the logical form output by the machine learning model as correct or incorrect based on the one or more raw bundling scores; and providing a confidence score for the logical form, the confidence score determined based on the classification of the logical form.

The present invention provides an apparatus for analyzing vulnerabilities of smart contract code, the apparatus including: a meta-information extraction unit configured to extract meta-information about input code, which is source code of a smart contract; an intermediate representation language conversion unit configured to convert the input code into an intermediate representation language based on the meta-information extracted by the meta-information extraction unit; a code analysis unit configured to detect whether there is vulnerable code or a rule violation case by analyzing the input code converted into the intermediate representation language by the intermediate representation language conversion unit; and a vulnerability detection unit configured to generate a final vulnerability model by detecting a vulnerability based on the vulnerable code or the rule violation case detected by the code analysis unit, and also provides an analysis method using the apparatus.

The invention discloses an ATS-oriented semantic-driven single-image-based 3D multi-target detection method, and the method comprises the steps: processing an input RGB image, and extracting a 3D bounding box of each object in the image; generating all potential 2D projections of the 3D object in the scene; extracting keywords, phrases and semantic information thereof in the description to form feature information Pt representing language description; fusing the 2D image information and the 3D geometric information of the object to obtain a complete object representation fa; correlating the language features extracted by the language description with the detected 3D object, and capturing a semantic correspondence relationship between the text and the visual modality; and filtering the targets according to the generated matching scores to obtain all targets according with natural language description. According to the method, the accuracy and efficiency of retrieval and recognition are improved, higher recognition precision is realized, the calculation complexity is remarkably reduced, the precision and speed of cross-modal retrieval can be improved, and the accurate recognition and positioning capability of traffic events is greatly improved.

Techniques are disclosed herein for resolving date / time expressions while transforming natural language to a logical form such as a meaning representation language. A class label for a token in a natural language utterance and a meaning representation for the natural language utterance can be predicted. The class label can be associated with a date / time expression. The meaning representation can include an operator and a value. When the value associated with the class label matches a predetermined value type or the operator matches a predetermined operator, the value and / or the operator can be modified, and an executable statement can be generated for the meaning representation. A query on a computing system can be executed using the executable statement.

The invention relates to a method for functional analysis of a source system (1), wherein a syntax tree model (3) is generated from the program code data of the source system (1). In order to propose a method for functional analysis of a source system (1), in particular with extensive program code data, which method makes it possible to better identify interrelated functional units and, with high result quality and at the same time low resource requirements, to improve further processing by means of a large language model, the invention provides that a syntax analysis unit (4) generates a syntax embedding model by decomposing the syntax tree model (3) into individual syntax element embeddings (5), each representing a syntax element, and stores said syntax embedding model in a syntax embedding model memory (6), and that a language analysis unit (8) generates a language embedding model by decomposing the program code data into individual language element embeddings (9), each representing a language element, and stores said language embedding model in a language embedding model memory (10), whereupon, by means of a query unit (12), at least one query embedding (13) is generated from a query (11) representing a functional element of the source system (1) and those syntax element embeddings (5) and language element embeddings (9) which lie within a predefined range of similarity to the query embedding (13) are retrieved from the syntax embedding model memory (6) and the language embedding model memory (10), whereupon the syntax elements represented by the retrieved syntax element embeddings (5) and the language elements represented by the retrieved language element embeddings (9) are transferred as result elements (14) to an output unit (15), which generates therefrom a functional description dataset (16) and outputs the latter.

The application discloses a 3D multi-target detection method based on semantic driving single image for ATS, which processes input RGB images, extracts 3D bounding boxes of each object in the images, generates all potential 2D projections of 3D objects in a scene, extracts keywords, phrases and semantic information in the description to form feature information P representing the language description t ; fuses 2D image information and 3D geometric information of the object to obtain complete object representation a ; associates language features extracted from the language description with the detected 3D object, captures the semantic correspondence between the text and the visual modal; filters the target according to the generated matching score to obtain all targets meeting the natural language description. The application improves the accuracy and efficiency of retrieval and recognition, realizes higher recognition accuracy, significantly reduces the computational complexity, can improve the accuracy and speed of cross-modal retrieval, and greatly improves the accurate recognition and positioning ability of traffic events.

Techniques are disclosed herein for calibrating confidence scores of a machine learning model trained to translate natural language to a meaning representation language. The techniques include obtaining one or more raw beam scores generated from one or more beam levels of a decoder of a machine learning model trained to translate natural language to a logical form, where each of the one or more raw beam scores is a conditional probability of a sub-tree determined by a heuristic search algorithm of the decoder at one of the one or more beam levels, classifying, by a calibration model, a logical form output by the machine learning model as correct or incorrect based on the one or more raw beam scores, and providing the logical form with a confidence score that is determined based on the classifying of the logical form.

The present invention provides an apparatus for analyzing vulnerabilities of smart contract code, the apparatus including: a meta-information extraction unit configured to extract meta-information about input code, which is source code of a smart contract; an intermediate representation language conversion unit configured to convert the input code into an intermediate representation language based on the meta-information extracted by the meta-information extraction unit; a code analysis unit configured to detect whether there is vulnerable code or a rule violation case by analyzing the input code converted into the intermediate representation language by the intermediate representation language conversion unit; and a vulnerability detection unit configured to generate a final vulnerability model by detecting a vulnerability based on the vulnerable code or the rule violation case detected by the code analysis unit, and also provides an analysis method using the apparatus.

Techniques for augmenting training data include accessing training data comprising a plurality of training examples comprising a first training example comprising a first natural language utterance and a first logical form for the first natural language utterance. A second natural language utterance is generated by adding or replacing one or more values in the first natural language utterance. A logical form for the second natural language utterance is generated. A second training example is generated, comprising the second natural language utterance and the logical form for the second natural language utterance. The training data is augmented by adding the second training example to the plurality of training examples to generate an augmented training data set. A machine learning model is trained to generate logical forms for utterances using the augmented training data set.

The present invention discloses a method, apparatus, device, and storage medium for identifying intent inconsistencies in distributed applications, aiming to address the technical issues of poor universality and recognition effectiveness of existing intent consistency identification methods. The method comprises the following steps: obtaining source code and annotations of a distributed application; generating annotation prompts for the annotations; extracting code features from the source code using an intermediate representation language; converting the annotation prompts into annotation embeddings and converting code features into code embeddings using a pre-trained embedding model; and determining intent consistency of the distributed application based on the annotation embeddings and code embeddings.

Techniques are disclosed herein for using named entity recognition to resolve entity expression while transforming natural language to a meaning representation language. In one aspect, a method includes accessing natural language text, predicting, by a first machine learning model, a class label for a token in the natural language text, predicting, by a second machine-learning model, operators for a meaning representation language and a value or value span for each attribute of the operators, in response to determining that the value or value span for a particular attribute matches the class label, converting a portion of the natural language text for the value or value span into a resolved format, and outputting syntax for the meaning representation language. The syntax comprises the operators with the portion of the natural language text for the value or value span in the resolved format.

In various examples, generating synthetic voices for speech for conversational systems and applications is described herein. Systems and methods described herein may generate data, such as data representing speaker embeddings (e.g., timbre, etc.) and / or frequency values (e.g., pitch, etc.), which is then used to generate audio data representing speech in synthetically produced voices. For instance, speaker embeddings may be used to generate a new speaker embedding associated with a synthetically produced voice, such as by linearly interpolating between the speaker embeddings and / or sampling an embedding space associated with speaker embeddings. Additionally, a frequency value associated with the synthetically produced voice may be identified, such as by randomly sampling from a distribution of frequency values. A component may then use the speaker embedding, the frequency value, and / or input data representing linguistic content to generate audio data representing the speech in the synthetically produced voice.

Techniques are disclosed herein for addressing catastrophic forgetting and over-generalization while training a model to transform natural language to a logical form such as a meaning representation language. The techniques include accessing training data comprising natural language examples, augmenting the training data to generate expanded training data, training a machine learning model on the expanded training data, and providing the trained machine learning model. The augmenting includes (i) generating contrastive examples by revising natural language of examples identified to have caused regression during training of a machine learning model with the training data, (ii) generating alternative examples by modifying operators of examples identified within the training data that belong to a concept that exhibits bias, or (iii) a combination of (i) and (ii).

To provide an apparatus and a method that can simply and efficiently analyze vulnerabilities that may occur in smart contract-based software operating on a blockchain.SOLUTION: An apparatus 100 for analyzing vulnerabilities of a smart contract code includes: a meta-information extraction unit 10 configured to extract meta-information about an input code, which is a source code of a smart contract; an intermediate representation language conversion unit 20 configured to convert the input code into an intermediate representation language on the basis of the meta-information extracted by the meta-information extraction unit 10; a code analysis unit 30 configured to detect whether there are a vulnerable code and a rule violation case by analyzing the input code converted into the intermediate representation language by the intermediate representation language conversion unit 20; and a vulnerability detection unit 40 configured to generate a final vulnerability model by detecting a vulnerability on the basis of the vulnerable code and the rule violation case detected by the code analysis unit 30.SELECTED DRAWING: Figure 1

The invention provides a user feedback detection method, which comprises the following steps of: modeling preset language information according to a causal architecture to obtain a structural causal graph; the structural causal diagram is represented by the following formula. Wherein semantic content factors in the causal architecture, semantic style factors in the causal architecture, text information included in the language information and label information included in the language information are expressed. And performing content focusing on the language information according to a preset large language model to obtain a focusing representation. And modeling according to the focusing representation and the structural causal graph to obtain a causal decoupling model. And detecting the user feedback according to the causal decoupling model to obtain a detection result. Therefore, accurate recognition of user feedback by the language model in dialect, slang, non-standard sentence pattern and other multi-style scenes is realized.

The invention relates to a high-level language compiling method and device, electronic equipment and a storage medium. The method comprises the steps that a high-level language code is analyzed into an abstract syntax tree; semantic parsing operation is executed on the abstract syntax tree to obtain a first language, and the first language is a high-level intermediate representation language; generating a second language based on the first language, wherein the second language is a low-level intermediate representation language; and generating a target file according to the second language. According to the embodiment of the invention, the abstract syntax tree is converted into the advanced intermediate representation language, so that software engineering design is more met, and a code structure is clearer.

Techniques are disclosed herein for calibrating confidence scores of a machine learning model trained to translate natural language to a meaning representation language. The techniques include obtaining one or more raw beam scores generated from one or more beam levels of a decoder of a machine learning model trained to translate natural language to a logical form, where each of the one or more raw beam scores is a conditional probability of a sub-tree determined by a heuristic search algorithm of the decoder at one of the one or more beam levels, classifying, by a calibration model, a logical form output by the machine learning model as correct or incorrect based on the one or more raw beam scores, and providing the logical form with a confidence score that is determined based on the classifying of the logical form.