136 results about "Word target" patented technology

The invention relates to the technical field of artificial intelligence, finance and medical health, and provides a retrieval method, device and equipment based on attention guidance and a medium. Training data is constructed according to the correlation between each specified lexical element and a retrieval question and a retrieval answer and the attention feature vector of each specified lexical element; a logistic regression classifier is trained to obtain a detector, and the light-weight detector can be used for recognizing utilized lexical elements in real time in the reasoning process, so that the higher-degree explanation of the generation process is realized; when the retrieval model is used for retrieval, the detector is used for detecting the target lexical elements which are being used in real time, the original probability distribution is corrected according to the utilization probability of each target lexical element so as to control the retrieval model to output the retrieval result, and the output probability of the used lexical elements is dynamically enhanced. Therefore, the attention mechanism of the guide model is dynamically focused on the context information highly related to the user query, and the context illusion phenomenon is fundamentally reduced.

The invention provides a text determination method based on a multi-modal large model, a multi-modal large model training method and equipment, and relates to the technical field of artificial intelligence, in particular to the technical fields of computer vision, deep learning, natural language processing, large models and the like. According to the specific implementation scheme, a first lexical element sequence is determined through a coding network in a multi-modal large model, a target image and a task description text corresponding to the target image, the first lexical element sequence is input into a large language model in the multi-modal large model, and an output first lexical element feature sequence is obtained, and inputting the last lexical element feature in the first lexical element feature sequence into each decoding network in the multi-modal large model, obtaining a target lexical element output by each decoding network, and obtaining a target text according to the first lexical element sequence and the target lexical element.

The embodiment of the invention provides a multidimensional vectorized text similarity calculation method based on syntactic enhancement, which belongs to the technical field of data processing, and specifically comprises the following steps: step 1, carrying out word segmentation and syntactic analysis on an input text to obtain a word segmentation result and a corresponding component tag; step 2, performing vectorization representation and fusion on word segmentation results by adopting a WoBERT model and a self-attention mechanism of Transform respectively to obtain a fusion vector; 3, dimension reduction processing is conducted on the fusion vector through a deep integration converter, and a target word vector with the high expression ability is generated; step 4, for the target word vectors of the same component labels, calculating a similarity value through a seme similarity formula; and step 5, based on all similarity values, adjusting a final similarity score through weighted average. According to the scheme, a text similarity matching method based on syntax and word vector combination is introduced, and the accuracy and efficiency of similarity calculation are improved.

The invention provides a text reinforcement learning method and device, electronic equipment and a computer storage medium, relates to the technical field of text reinforcement learning, and is applied to a text generation model. The method comprises the following steps: determining one or more target vocabularies in a statement to be analyzed, and generating a replacement word set according to the target vocabularies; replacing the target vocabulary in the to-be-analyzed statement with a current candidate replacement word to obtain a replacement statement; calculating one or more difference index values between the replacement statement and the to-be-analyzed statement; wherein the difference index value is used for quantifying the influence of the replacement of the target vocabulary on the original sentence from different dimensions; based on the difference index value, determining an importance weight of the target vocabulary in the to-be-analyzed statement; wherein the importance weight is used for redistributing the reward of the single character in the sentence. Vocabulary importance is quantified through disturbance analysis and multi-dimensional evaluation, and then fine-grained distribution of rewards is achieved.

According to the method and device for generating the fine-grained semantic description from the action video data, the training data set is established based on the isolated word sign language recognition data set and the continuous sign language recognition data set containing the word target notation, and the action video data and the action description text data of fine-grained semantic description modeling are obtained; a fine-grained semantic action description style pre-training generation model is obtained through a training framework comprising an action video feature coding module, a multi-modal feature fusion module and a text feature coding module by combining user cue words and system cue words and introducing a mask reconstruction mechanism, action video data is adopted for fine tuning, a loss function is established, and a fine-grained semantic action description style is obtained. The fine-grained semantic action description generation model is used for generating high-quality fine-grained semantic action description data, and the problem that the current fine-grained semantic action description data is insufficient is solved. And the stability and the accuracy of a generated result are ensured when high-dynamic complex scenes such as sign language videos and interactive actions are processed.

The invention provides a conversation system interaction method and device based on KV cache multiplexing, and the method comprises the steps: determining a reference cue word when a target cue word is received, and then determining an editing operation set corresponding to the minimum number of editing times needed for converting the reference cue word into the target cue word, and for each decoder layer, determining a unified attention map of the reference cue word in each decoder layer, and for each decoder layer, determining a target lexical element of the target cue word needing to recalculate the KV cache in the decoder layer according to the editing operation set and the unified attention map of the reference cue word in the decoder layer. According to the method, the target lexical elements are determined by editing the operation set and the unified attention graph, so that the target lexical elements, needing to be recalculated in the KV cache, of the target cue word in each decoder layer can be accurately recognized, and the KV cache of the reference cue word can be reused for other lexical elements, in each decoder layer, of the target cue word; the reuse rate of the KV cache and the response speed of the dialogue system are improved as much as possible.

The invention provides a speculation sampling method for a large language model, and the method comprises the steps: determining a target topic related to a target cue word from a plurality of preset topics after obtaining the target cue word of the input large language model; each target topic is provided with a corresponding word list, and after the target topic is determined, speculation sampling based on the frequency of each text fragment recorded in the word list can be carried out based on the target word list corresponding to the target topic. Therefore, prediction of the decoding stage of the large language model can be carried out by using a speculation sampling result. The speculation sampling based on the word list can reduce the occupation of video memory resources. And word lists of different topics are distinguished, so that the speculation sampling accuracy is improved.

The invention relates to the field of intelligent analysis, and provides an information interaction method for joint simulation of a main network and an active distribution network, and the method comprises the steps: extracting a first to-be-evaluated key information item from a key information library, carrying out the word segmentation of a specific target through a deep learning model based on natural language processing, and obtaining a first to-be-evaluated key information item; and performing semantic embedding coding on each specific target word granularity after word segmentation and the first to-be-evaluated key information item, so as to obtain a semantic query result according to comparison between a semantic query matching coefficient between a semantic feature of each specific target word granularity and a semantic feature of the first to-be-evaluated key information item and a preset threshold value. And whether the first to-be-evaluated key information item is included is automatically determined. Therefore, the key information items can be automatically screened, the time and energy required by manual screening are reduced, the information processing speed and efficiency are improved, and meanwhile the response speed of the system is improved.

The invention provides a user intention mining processing method and device and related equipment, and aims to solve the problem of low user intention mining efficiency when self-service voice service intelligent equipment constructs a dialogue system in the traditional technology. The user intention mining processing method comprises the following steps: obtaining an initial to-be-mined text of a to-be-mined user intention, and according to a TF-IDF value, obtaining an initial to-be-mined text of the to-be-mined user intention; the method comprises the steps of obtaining word weights of corresponding lexical units, filtering lexical unit sentence texts corresponding to the lexical units according to the word weights to obtain target lexical unit sentence texts, vectorizing the target lexical unit sentence texts, determining target lexical unit sentence text vectors, clustering all the target lexical unit sentence text vectors, and determining target user intentions according to clusters. According to the user intention mining method and device, repeated corpora and corpora low in importance degree can be removed by filtering the lexical text corresponding to the lexical, the data scale is reduced, the user intention mining efficiency and quality are improved, the user intention mining effect is improved, and the user intention mining method and device can be applied to the field of medical health.

The invention relates to the technical field of watermarks, in particular to a watermark generation method and device based on a large model, equipment and a medium. The method comprises the following steps: acquiring a target prompt word sequence, and acquiring a word sequence to be generated according to the target prompt word sequence; selecting one or more words from the word sequence as target words according to a preset random selection seed, and embedding watermark information into each target word according to preset watermark intensity to obtain each watermark mark word; according to the sequence of the words, splicing the watermark marking words and the words except the watermark marking words to obtain a watermark generation text; and generating a target watermark text according to the target prompt word sequence and the watermark generation text. According to the method provided by the invention, the watermark can be embedded on the premise of not influencing the quality of the model generated text as much as possible, and the watermark generated text can be detected relatively quickly.

The invention discloses a speculation decoding method and device based on a large language model, equipment and a medium, and relates to the technical field of large models, and the method comprises the steps: determining the output probability distribution of a plurality of lexical element positions, and determining a confidence score based on the output probability distribution; determining a target control parameter according to the target state signal, performing aggregation calculation on the confidence coefficient vector to obtain an aggregation confidence coefficient, and determining a target draft length through the aggregation confidence coefficient and the target control parameter; determining a verification tolerance threshold based on the target control parameter and the confidence score, generating a draft sequence based on the target draft length by using a target deep learning model, and verifying each lexical element in the draft sequence based on a preset large language model and the verification tolerance threshold; and adding the successfully verified target lexical units to the end of the currently generated sequence to obtain a target sequence, determining the target sequence as the currently generated sequence, and skipping again to perform the next round of speculative decoding. According to the invention, the efficiency of speculative decoding can be improved.

The embodiment of the invention discloses a request processing method and device for a large language model. The method is implemented on the basis of a plurality of generation sections preset for an output text of the large language model, and each generation section is configured with a corresponding start lexical element, a corresponding termination lexical element and a corresponding query sampling corpus. The method comprises the following steps: firstly, aiming at a first request, under the condition that the first request is judged to be in a decoding stage, determining a corresponding first generation section according to a first lexical element sequence included in the first request; and then, based on a first query sampling corpus corresponding to the first generation section and the large language model, performing speculation sampling processing aiming at the first lexical element sequence to obtain a target lexical element sequence after the first lexical element sequence, and taking the target lexical element sequence as a processing result of the first request.

The embodiment of the invention provides a text generation method and system and a model training method and system. According to the text generation method, the text generation system can obtain a demand target prompt text for representing text generation and input the target prompt text into a trained diffusion model so as to generate a target text through the diffusion model. Wherein the target text comprises a plurality of lexical elements. The diffusion model is trained to have the ability to predict the noisy text for multiple rounds based on the target prompt text to gradually generate each lexical element in the target text. In each round of prediction, the diffusion model generates part of lexical elements in the target text in a parallel mode. Aiming at a target lexical element in the target text, the diffusion model takes forward text information and backward text information of the target lexical element as reference when the target lexical element is generated, the forward text information comprises text information located in front of the target lexical element, and the backward text information comprises text information located behind the target lexical element.

The embodiment of the invention provides an observation information generation method, device and equipment. According to the scheme, the method comprises the steps that in the process that a generative large language model generates reply information for a target request, a probability distribution vector of a target lexical element is collected; then, based on the probability distribution vector, determining recommendation lexical element information with high probability; acquiring sampling parameter information corresponding to the target request; afterwards, the recommended lexical element information, the sampling parameter information and the target lexical element are stored in an associated mode, and quality observation information is obtained so as to be used for analyzing reasons for quality abnormity of the reply information.

The invention discloses a text data classification method and device, electronic equipment and a nonvolatile storage medium. The method comprises the steps of obtaining target text data; adopting a first model to match the target text data with vocabularies in a target word bank, and determining a first score corresponding to the target text data according to a matching result, a second model is adopted, the first semantic feature corresponding to the target text data is matched with a target semantic feature library, a second score corresponding to the target text data is determined according to the matching result, and the operand and complexity of the second model are larger than those of the first model; and determining a classification result of the target text data according to the first score and the second score. According to the method and the device, the technical problem of poor text data classification effect caused by limitation of a single model on semantic comprehension and classification decision in related technologies is solved.

The invention provides a coal industry lexicon classification method, which comprises the following steps: acquiring an initial lexicon, a long sequence text and a short sequence text, inputting the initial lexicon and the long sequence text into a first text classification model, acquiring first prediction classification information, and inputting the first prediction classification information into a second text classification model, inputting the long-sequence text into a first text classification model to obtain second prediction classification information, inputting the initial word bank and the short-sequence text into a second text classification model to obtain third prediction classification information, inputting the short-sequence text into the second text classification model to obtain fourth prediction classification information, and determining the target lexicon according to the first annotation classification information, the second annotation classification information, the first prediction classification information, the second prediction classification information, the third prediction classification information and the fourth prediction classification information, thereby improving the classification effect of the lexicon in the coal industry.

The invention relates to the technical field of artificial intelligence, is suitable for the financial field and the medical health field, and provides a verbal skill generation method and device, equipment and a storage medium, and the method comprises the steps: obtaining a user input text of which a topic deviates from a business topic, and extracting a target word group from the user input text; determining a first response template text from a plurality of preset response template texts associated with different business themes; generating a second response template text based on the target word group and the first response template text; analyzing the user input text and the second response template by using a preset language model to generate a plurality of target lexical elements for responding to the user input text, and obtaining probability data corresponding to each target lexical element; and based on the multiple target lexical elements and the probability data corresponding to each target lexical element, generating a target response verbal skill used for guiding the topic to return to the business theme. The method can help an artificial intelligence customer service to naturally guide the dialogue to return to the business theme.

This application relates to the field of artificial intelligence security technology and discloses a method, apparatus, device, and medium for security optimization of a large language model, comprising: acquiring an attack task dataset; wherein the attack task dataset includes at least one attack task type and multiple malicious instruction data under the attack task type; generating a general word-level adversarial suffix corresponding to the attack task type based on the malicious instruction data, and generating a target word-level adversarial suffix corresponding to the malicious instruction data based on the general word-level adversarial suffix; performing a first fine-tuning on the large language model based on the target word-level adversarial suffix to obtain a first fine-tuned model; and performing a second fine-tuning on the first fine-tuned model based on the acquired target semantic-level adversarial hints to obtain a second fine-tuned model. This application can improve the ability of a large language model to resist various known and unknown forms of jailbreak attacks.

This application discloses a vocabulary prediction method, apparatus, storage medium, and computer device, wherein the method comprises: obtaining a target text for a target user, obtaining each first target vocabulary in the target text and the order relationship between each first target vocabulary; using a grade evaluation model to obtain a target text grade of the target text based on each first target vocabulary and the order relationship between each first target vocabulary; using a vocabulary prediction model to obtain a text vocabulary corresponding to the target text based on each first target vocabulary and the vocabulary grade corresponding to each first target vocabulary; and determining a predicted vocabulary of the target user based on the target grade vocabulary grade and the text vocabulary. Using this application, the accuracy of vocabulary prediction is improved.

This application provides a text processing method, apparatus, device, and storage medium. The method involves segmenting the text to obtain several words to be grouped; calculating a first semantic similarity between the words to be grouped; grouping the words to be grouped into word combinations based on the first semantic similarity, with each word combination containing words to be grouped with similar semantics; searching for words in the word combinations that do not match a preset correct word as replacement words; replacing the replacement words with the preset correct word to obtain a target word combination; and using the target word combination to correct the word combinations in the text to obtain the target text. This application, based on the first semantic similarity, automatically performs consistency checks and processing on the text in the front-end system, improving the efficiency and accuracy of the front-end system's text consistency checks, thereby saving manpower and time costs.

Embodiments of the present application provide a cross-model collaborative reasoning method and device, equipment and a storage medium, relating to the technical field of intelligent computing equipment. The input prompt is sent to the first model for reasoning to generate a probability value corresponding to each word element at the current time step. The highest probability value is determined from the probability value, and the marginal benefit of the current time step is determined according to the highest probability value. The current benefit threshold corresponding to the current time step is obtained, and the decision variable is determined according to the marginal benefit and the current benefit threshold. The target word of the current time step is obtained based on the decision variable, and the output sequence corresponding to the input prompt is obtained according to the target word sequence before the current time step and the target word. The on-demand switching of the double model is realized by the dynamic determination of the marginal benefit and the benefit threshold. The first model is used to realize low-energy-consumption reasoning at the time step with high reasoning confidence. The second model is used to ensure accuracy at the key time step with low reasoning confidence. The reasoning accuracy is ensured while reducing the reasoning energy consumption.

The application discloses a semantic recognition model based on NPL and a training method thereof, relates to the technical field of semantic recognition model training, and aims to solve the problem of low recognition accuracy of a semantic recognition model. The current whole sentence is searched, multiple segmented sentences are screened out, words in the segmented sentences are recognized, the part-of-speech information of the words is collected, the part-of-speech weight under the corresponding part-of-speech information is called, the number of repeated occurrences of each word in the current whole sentence is counted and the repeated word frequency is calculated, the processing level of each word is calculated by comprehensively considering the repeated word frequency and the part-of-speech weight, the target words are screened out and recognized, the semantics of the target words are acquired and the number of semantics is counted under the current interactive window, the interactive duration of the target words is analyzed by calling historical data, the number of word reference sources is set as a reference, the number of semantics of the target words and the interactive duration are comprehensively considered to adjust the number of word reference sources of each target word, and the semantic analysis accuracy of each target word in the interactive sentence is improved.

The invention discloses a two-stage new word discovery method and device, equipment and a storage medium, and belongs to the field of natural language process.The method comprises the steps that a character field set containing a plurality of to-be-verified target character fields is extracted from a corpus text, and the adjacent entropy of each target character field in the character field set is determined; determining a word set of the corpus text according to the adjacent entropy of each target character field in the character field set; and inputting the words in the word set into a new word discovery model, so as to determine a first target word from the word set according to the probability feature of each word determined by the new word discovery model. Based on the method provided by the embodiment of the invention, the problems confronted based on rules and statistical methods in the prior art are overcome, the problems of relying on high-quality expert knowledge and pre-screening a large amount of data are solved, the efficiency and accuracy of new word discovery are improved, and the problems of poor universality and high maintenance cost in the existing new word discovery process are solved.

The present invention relates to the field of artificial intelligence, and discloses a word retrieval method, device, equipment and storage medium based on binary mutual information, which are used to increase the accuracy of extracting keywords, and then retrieve effective message responses. The method of the present invention includes: obtaining a target question text sent by a target user; segmenting the target question text to obtain multiple candidate words, each candidate word is unique; calling a preset corpus to determine the word frequency TF and inverse document frequency IDF of multiple target words in multiple candidate words; calculating the binary mutual information of each target word in multiple target words according to a preset formula; obtaining the adjustment factor of each target word in multiple target words according to a preset algorithm; calculating the weight value of each target word according to the adjustment factor of each target word and the binary mutual information of each target word; determining the keyword of the target question text according to the weight value of each target word, and retrieving the corresponding answer according to the keyword.

The invention relates to the field of data processing, in particular to a character ambiguity elimination method and system based on artificial intelligence. The method comprises the following steps: constructing a lexicon, and dividing all reference words into strong correlation words and weak correlation words of target words; removing unintentional words in the answer sentences to obtain effective answer sentences, calculating a distribution coefficient of any word in the effective answer sentences, and constructing word vectors of the effective answer sentences according to the distribution coefficient; clustering the word vectors of all the effective answers to obtain a plurality of clusters, traversing any cluster to obtain reciprocals of the number of the strongly related words in the marked question corresponding to each effective answer, and taking a mean value of all the reciprocals as an ambiguous value of the cluster; and traversing to obtain an ambiguity value of each cluster, constructing an ambiguity value sequence of the contact word to the target word, and calculating ambiguity evaluation to complete ambiguity elimination. According to the technical scheme, the accuracy of text understanding can be improved, and recognition errors caused by character ambiguity are reduced.

Disclosed are a word vector generation method and device. The word vector generation method includes: obtaining corpus data including at least two text units, each text unit including at least one word; determining the weight of each text unit in the corpus data based on the distribution of each group word in a group word set and each target word in a target word set in the text units of the corpus data; determining sample data for training a word vector model based on the corpus data and the weight of each text unit in the corpus data; training the word vector model using the sample data, and obtaining the word vector of at least one word in at least one text unit in the corpus data from the trained word vector model.

The application discloses a cross-media model error attribution and auxiliary correction method based on causal decoupling, which comprises the following steps: image segmentation is performed on an input image, and the input image is divided into a plurality of sub-regions; a target function is solved according to the sub-regions, the minimum key region which supports the generation of a target word element to the greatest extent and removes the generated ability is sorted according to the contribution to the generation decision, and an attribution saliency map is generated; an influence score is calculated according to the change of the generation probability of the target word element in the process of inserting the sub-regions in the ordered subset; the ordered subset, the attribution saliency map and the influence score are evaluated from the dimensions of fidelity, positioning ability and error correction guiding ability, and an evaluation result is obtained; the evaluation result is obtained without relying on the internal gradient, attention weight or activation map of the model, the dependence on the internal structure of the model is reduced, the regional attribution of an arbitrary word element set is realized, and the relative dependence of the generation process on visual evidence and language prior is quantified.

Embodiments of the present specification disclose a key information extraction and live content processing method, device and equipment. The scheme comprises: obtaining original text to be extracted key information; determining an attention weight vector for the original text, and correspondingly generating a first target word according to the attention weight vector; judging whether the first target word is included in the original text; if yes, obtaining the attention weight of the first target word when generating the first target word according to the attention weight vector, generating penalty data for the first target word according to the attention weight of the first target word, and continuing to generate a second target word according to the penalty data, wherein the penalty data is used to reduce the possibility that the second target word generated continues to be the same as the first target word; and generating key information according to the first target word and the second target word.

The invention provides a code generation method and device, electronic equipment and a storage medium. The method comprises the following steps: analyzing at least one piece of key information in demand information corresponding to a natural language search text; in order to determine the importance degree of the key information, respectively generating second probability distribution obtained according to semantic features of the natural language search text and first probability distribution corresponding to a preset lexical element library obtained according to mask semantic features, since the mask semantic feature is the semantic feature corresponding to the natural language search text after hiding the key information, the influence degree of the key information on the probability distribution of the generated lexical elements can be determined according to the difference distribution between the first probability distribution and the second probability distribution. According to the method, the difference value distribution and the second probability distribution are obtained, the target lexical element sequence is determined according to the difference value distribution and the second probability distribution, the accuracy of the target lexical element sequence is improved, finally, the target code is generated according to the target lexical element sequence, and the matching degree of the target code and the demand information corresponding to the natural language search text is improved.

The present application relates to the field of artificial intelligence, and particularly to a method, device, computer device, storage medium and computer program product for entity information processing, which can be used in the field of financial technology and can also be used in any field other than the field of financial technology. The method includes: obtaining the corpus to be processed; constructing a lexical co-occurrence graph with each word in the corpus to be processed as a node and the co-occurrence relationship between any two words as the edge of the node; the edge has a corresponding weight; the weight is used to represent the co-occurrence times of the words corresponding to the nodes connected by the edge in the corpus to be processed; inputting the lexical co-occurrence graph into a target graph convolutional neural network to obtain the target word vectors corresponding to each word; determining the target words belonging to the same entity according to the target word vectors corresponding to each word; the similarity between the target word vectors corresponding to the target words is greater than a preset similarity threshold. By using this method, the target words belonging to the same entity can be accurately determined.