Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, are provided. One example method includes receiving an input prompt. For a first token in the input prompt, m first utility scores are determined, where the kth first utility score (k=1, 2, . . . , m) indicates a relevance of the kth token in a vocabulary of m tokens with respect to the first token; the vocabulary of m tokens are assigned to buckets based on m first utility scores; m second utility scores are determined, where the kth second utility score is determined based on the kth first utility score, a quantity of tokens assigned to the same bucket as the kth token, and a noise; a candidate set of tokens is determined based on m second utility scores; and a replacement token is sampled from the candidate set. The method further includes generating a perturbed prompt comprising the replacement token.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, are provided. One example method includes receiving an input prompt in a natural language, the input prompt comprising a sequence of tokens. For a first token in the sequence of tokens, a context metric of the first token is determined in the context of the sequence of tokens in the input prompt, a distance metric is determined to indicate a distance between an embedding of the first token and an embedding of each of a set of tokens included in a vocabulary, a utility metric is determined based on the context metric and the distance metric, and a replacement token of the first token is sampled from the vocabulary based on the utility metric. The method further includes generating a perturbed prompt corresponding to the input prompt, the perturbed prompt comprising the replacement token.

Abstract: An air guide ring includes an annular air guide portion having an air duct that accommodates an impeller of an axial flow fan configured to rotate therein. The annular air guide portion has an inner side wall and an outer side wall, the air duct is enclosed by the inner side wall, and a plurality of noise reduction devices is arranged on the inner side wall of the annular air guide portion in a circumferential direction of the air guide ring The plurality of noise reduction devices is configured to reduce noise in the air duct.

Abstract: The present disclosure belongs to the field of image processing technologies, and in particular, to an image decoding method and apparatus, an image coding method and apparatus, a device and a storage medium. The image decoding method of the present disclosure includes: extracting image residual data or extended residual data from an image bitstream, and obtaining a plurality of extended residual groups based on the extracted image residual data or extended residual data; obtaining respective image reconstruction features corresponding to the extended residual groups by performing residual restoration on each of the plurality of extended residual groups; obtaining reconstructed feature data by performing spatial resolution amplification processing on the respective image reconstruction features corresponding to the extended residual groups; and obtaining a reconstructed image block by performing image reconstruction according to the reconstructed feature data.

Abstract: A falling film evaporator includes a housing defining an accommodating cavity and a cover provided in the accommodating cavity. The cover defines a cover accommodating cavity having a lower cover opening. The falling film evaporator further includes a gas flow channel located between the housing and the cover and enabling the lower cover opening to be in communication with a gas outlet provided at an upper portion of the accommodating cavity, and a plurality of flow guide members arranged in the gas flow channel along a height direction of the housing, where, in the height direction, one of two adjacent flow guide members is connected to the housing and is separated from the cover by a first distance, and the other of the two adjacent flow guide members is connected to the cover and is separated from the housing by a second distance.

Abstract: The present application belongs to the technical field of image coding and discloses an image reconstruction method and apparatus, device, and storage medium. The image reconstruction method includes: acquiring a quantization parameter and a prediction mode corresponding to a current coding block; adjusting a prediction granularity corresponding to the prediction mode; predicting according to the adjusted prediction granularity to obtain a prediction sample block; obtaining a reconstructed sample block based on the prediction sample block and the quantization parameter.

Abstract: A lithium ion secondary battery negative electrode additive includes polyaspartate salt and water.

Abstract: A display panel and a display device are provided. The display panel includes a pixel driving circuit on a side of a substrate. The pixel driving circuit includes a drivier transistor; a ninth transistor with a first electrode being connected to a third initial signal line, and a second electrode being connected to a first electrode of the driver transistor; an eighth transistor with a first electrode being connected to a gate of the driver transistor; a first transistor with a first electrode being connected to a first initial signal line, and a second electrode being connected to a second electrode of the eighth transistor; and a second transistor with a first electrode being connected to the second electrode of the eighth transistor, and a second electrode being connected to a second electrode of the driver transistor.

Abstract: The present disclosure provides a decoding method and apparatus, a coding method and apparatus, and a device. The method comprises: on the basis of on a first code stream corresponding to a current block, acquiring a target probability distribution model corresponding to a first scaling factor; on the basis of the target probability distribution model corresponding to the first scaling factor, decoding a second code stream corresponding to the current block, to obtain a corrected residual feature corresponding to the current block; and, on the basis of the corrected residual feature, determining a reconstructed image block corresponding to the current block. The present disclosure improves the coding performance and the decoding performance.

Abstract: The current disclosure provides sensitive diagnostic methods that do not require invasive angiography. It was found that TNFAIP2 was secreted into the sera of subjects with atherosclerosis. This novel biomarker can be used to detect early-stage atherosclerosis. Accordingly, aspects of the disclosure relate to a method for measuring a TNFAIP2 level comprising: obtaining a biological sample from a subject having, suspected of having, or at risk of having atherosclerosis; and measuring the TNFAIP2 level in a biological sample from the subject as compared to a control. Also disclosed is a method for treating or preventing atherosclerosis in a subject, the method comprising administering an effective amount of a treatment for atherosclerosis to the subject after measuring the TNFAIP2 level in a biological sample from the subject as compared to a control.

Abstract: Example computer-implemented methods and systems for secure random noise generation are disclosed. One example method includes generating, by a first party, n random first bits and n -bit first strings. The first party generates, based on the n -bit first strings and the n random first bits, n pairs of -bit input messages. The first party receives n pairs of -bit second strings. The first party performs n 1-out-of-2 random oblivious transfers (ROTs) of the n pairs of -bit input messages from the first party to a second party. The first party generates, based on the n -bit first strings, a first random number. The first party performs, based on the first random number, secure multiparty computation (MPC) that involves the first party and the second party.

Abstract: A pixel circuit, a driving method and a display device. The pixel circuit includes a driving circuit, a first control circuit, a compensation control circuit and a first initialization circuit; the first control circuit is configured to control to connect the control end of the driving circuit and the connection node under the control of a first scan signal; the compensation control circuit is configured to control to connect the connection node and the first end of the driving circuit under the control of a second scan signal; the first initialization circuit is configured to write a first initialization voltage into the connection node under the control of an initialization control signal; the driving circuit is configured to control to connect the first end and a second end of the driving circuit under the control of a potential of the control end thereof.

Abstract: Embodiments of this specification provide a data processing method, apparatus, and device. The method includes: when it is detected that there is a feature data missing in newly added feature data in first data, obtaining second data corresponding to the first data; inputting historical feature data in the first data into a pre-trained encoder model to obtain a first compression score corresponding to the historical feature data in the first data, and inputting historical feature data in the second data into the pre-trained encoder model to obtain a second compression score corresponding to the historical feature data in the second data; and determining target data corresponding to the first data in the second data based on the first compression score and the second compression score, and performing imputation processing on the newly added feature data in the first data based on newly added feature data in the target data.

Abstract: Embodiment of the disclosure provides method, apparatus, device and storage medium for data query. In the method, first data is obtained by a first device based on a request for data query. The request indicates a first data attribute to be classified and aggregated and at least a second data attribute and a third data attribute for classification. The first data at least includes data associated with a second data attribute. The third data fragment of a result of data query is generated based on the first data fragment of the first data and the second data fragment of the second data received from the second device. The second data at least includes data associated with the third data attribute. The third data fragment includes an element corresponding to the first data attribute classified and aggregated based on at least the second data attribute and the third data attribute.

Abstract: A pixel driving circuit includes: a first reset module, connected to a first node, a first initialization signal terminal and a reset signal terminal, and configured to write a signal of the first initialization signal terminal into the first node in response to a signal of the reset signal terminal; a driving module, connected to the first node, a second node and a third node, and providing a driving current to the third node based on voltages of the first node and the second node; a coupling module, connected to the first node and a preset node, and configured to couple a voltage change of the first node to the preset node and couple a voltage change of the preset node to the first node; and a first switch module, connected to a constant voltage terminal, the preset node and a switch signal terminal.

Abstract: A computer-implemented method includes: mapping a first group of user identifiers that are active in a first region to generate a first group of anonymized user nodes of the first region; communicating with one or more second devices of a second region, using a privacy set intersection (PSI) protocol, to identify one or more intersecting signals accessed by the first group of anonymized user nodes of the first region and a second group of anonymized user nodes of the second region; constructing a user-signal-user graph based on the one or more intersecting signals and the first group of anonymized user nodes of the first region; communicating with the one or more second devices of the second region via secret sharing to calculate a weight of each intersecting signal; and constructing a use-user graph with at least one edge determined by the weight of each intersecting signal from the user-signal-user graph.

Abstract: An apparatus for forming electrical connections comprises a metal transition conductor, a terminal portion of a first electrical cable electrically and mechanically coupled to the metal transition conductor, a terminal portion of a second electrical cable electrically and mechanically coupled to the metal transition conductor, and a terminal portion of a third electrical cable electrically and mechanically coupled to the metal transition conductor. Each of the first, second, and third electrical cables terminates at the metal transition conductor. At least two of the first, second, and third electrical cables comprise conductors of different metal materials, and at least two of the first, second, and third electrical cables comprise conductors of different wire gauges.

Abstract: Generation of instructional interactions in a virtual session includes receiving first environment data associated with a first user device and a second environment data associated with a second user device. The first environment data and the second environment data are analyzed to determine the first environment is different from the second environment data. A first set of instructional interactions associated with at least one of the first user device or a first entity is detected. A second set of instructional interactions associated with at least one of the second user device or a second entity is generated based on the application of the language model on the first set of instructional interactions, the first environment data, and the second environment data. The generated second set of instructional interactions is outputted.

Abstract: A system for refrigerant leakage detection includes one or more sensors configured to detect one or more parameters of a building system including a refrigerant. The system further includes one or more storage devices storing instructions thereon that, when executed by one or more processors, cause the one or more processors to receive sensor data from the one or more sensors; apply the sensor data to a long short-term memory (LSTM) model to generate predicted sensor data corresponding to the one or more sensors; receive subsequent sensor data from the one or more sensors; compare the predicted sensor data to the subsequent sensor data; determine that the building system has a refrigerant leakage based on the comparison of the predicted sensor data to the subsequent sensor data; and, responsive to determining that the building system has the refrigerant leakage, take an action to address the refrigerant leakage.

Abstract: A pixel circuit, a driving method and a display device. The pixel circuit includes a driving circuit, a first initialization circuit and a reset circuit; the first initialization circuit configured to write a first initial voltage into the first end of the driving circuit under the control of an initialization control signal; the reset circuit is configured to write a reset voltage into the second end of the driving circuit or the first end of the driving circuit under the control of a second scan signal; the driving circuit is configured to control to connect the first end of the driving circuit and the second end of the driving circuit under the control of a potential of a control end of the driving circuit.