Abstract: A method, apparatus, device, and medium for managing a contrastive learning model are provided. In one method, in a first training phase, a first contrastive learning model is generated by training the contrastive learning model with a first training sample set, a negative sample pair of the first training sample set comprising only data segments from different data sequences. In a second training phase, a second contrastive learning model is generated by training the first contrastive learning model with a second training sample set, a negative sample pair in the second training sample set comprising data segments from the same data sequence. Knowledge in terms of the appearance of the samples can be fully obtained in the first training phase, and knowledge in terms of the appearance and dynamics of the samples can be fully obtained in the second training, with example implementations of the present disclosure.

Abstract: A solution for generating a positive sample pair of a contrastive learning model is provided. In one method, a first data segment and a second data segment are respectively obtained from a first data sequence in plurality of data sequences for training the contrastive learning model, the first data sequence comprising a plurality of data frames. A data frame is selected from a second data sequence in the plurality of data sequences. A third data segment is generated based on the second data segment and the data frame. A positive sample pair for training the contrastive learning model is determined using the first data segment and the third data segment. In this way, positive samples that are more difficult to distinguish can be provided, thereby increasing the accuracy of the contrastive learning model and improving the training efficiency and accuracy of downstream models.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may receive, from a base station, a configuration that defines a quantity of additional guard resource elements (REs) in a frequency domain between different frequency division multiplexed demodulation reference signal (DMRS) code division multiplexing (CDM) groups. The additional guard REs in the frequency domain may not be associated with signal transmissions. The UE may perform a channel estimation based at least in part on the quantity of additional guard REs in the frequency domain between the different frequency division multiplexed DMRS CDM groups. Numerous other aspects are described.

Abstract: A solution for generating a negative sample pair of a contrastive learning model is provided. In one method, a first data segment is obtained from a first data sequence in a plurality of data sequences for training the contrastive learning model, and a second data segment is obtained from a second data sequence in the plurality of data sequences. A data frame is selected from a further data sequence than the second data sequence in the plurality of data sequences. A third data segment is generated based on the second data segment and the data frame. A negative sample pair for training the contrastive learning model is determined based on the first data segment and the third data segment. Therefore, richer semantic information can be introduced into negative sample pairs in terms of appearance, and further the accuracy of the contrastive learning model can be improved.

Abstract: Embodiments of a system, device and method for configuring reference signaling are disclosed. In some embodiments, a method includes determining, by a wireless communication device, N sets of parameters, determining, by the wireless communication device according to a layer 1 measurement, whether a condition of an event is satisfied, and identifying, by the wireless communication device according to the determination, M sets of parameters from the N sets of parameters. In some embodiments, N is a positive integer value. In some embodiments, M is smaller than or equal to N.

Abstract: The invention provides a power supply including at least one power output port, at least one status alert component, and at least one output port status monitoring module. The status alert component generates at least one visual prompt based on an alert signal. The output port status monitoring module includes at least one temperature sensor adjacent to the power output port, a microcontroller connected to the temperature sensor and sensing an output current from the power output port, and a reset signal generator connected to the microcontroller. The microcontroller comprises at least one port status alert condition that takes a temperature and the output current of the power output port as decision factors. The microcontroller outputs the alert signal to the status alert component when the port status alert condition is met and maintains the status until a reset signal provided by the reset signal generator is received.

Abstract: A coated substrate for an electronic device can include a substrate, a basecoat layer on the substrate, and an anti-fingerprint topcoat layer on the basecoat layer. The substrate can include a metal or metal alloy. The basecoat layer can include pigment particles and a first one-part thermally cured polymeric resin. The anti-fingerprint topcoat layer can include a second one-part thermally cured polymeric resin and an anti-fingerprint material. The anti-fingerprint material can include a fluoropolymer, a silane, or a combination thereof. The basecoat layer can be cured before applying the anti-fingerprint topcoat layer on the basecoat layer.

Abstract: A method of training and applying contrastive learning model. The method includes obtaining a sample set and label information for training contrastive learning model, the sample set including a plurality of first samples of a first modality and a plurality of second samples of a second modality, the label information indicating a correlation between samples of the plurality of first samples and samples of the plurality of second samples; determining whether sample mixing is to be performed on the first modality or the second modality; in accordance with a determination that sample mixing is to be performed on the first modality, generating at least one first mixed sample of the first modality by mixing at least one pair of first samples among the plurality of first samples; and training the contrastive learning model at least based on the at least one first mixed sample and first mixed label information.

Abstract: The present disclosure discloses an in situ U—Pb dating method for calcite, including: cutting a calcite sample to prepare an epoxy resin sample target; placing the sample in a laser ablation sample chamber, and adjusting a position of the sample in an optical axis direction; conducting line scanning ablation on the sample target, and measuring ion signal intensity data of 43Ca, 88Sr, 139La, and 238U; conducting two-dimensional (2D) element imaging to obtain a 2D element content distribution map; according to the 2D element content distribution map, determining a high-U analysis target area, conducting point ablation on the high-U target area, and measuring ion signal intensity data of 206Pb, 207Pb, and 238U; and after the element signal data is obtained, calculating 207Pb/206Pb and 238U/206Pb fractionation coefficients, correcting ratios of an unknown sample, constructing a Tera-Wasserbug diagram, and calculating age data and an initial Pb isotope (207Pb/206Pb) composition of the calcite sample.

Abstract: Provided is a system for detecting an anomaly in a multivariate time series that includes at least one processor programmed or configured to receive a dataset of a plurality of data instances, wherein each data instance comprises a time series of data points, determine a set of target data instances based on the dataset, determine a set of historical data instances based on the dataset, generate, based on the set of target data instances, a true value matrix, a true frequency matrix, and a true correlation matrix, generate a forecast value matrix, a forecast frequency matrix, and a forecast correlation matrix based on the set of target data instances and the set of historical data instances, determine an amount of forecasting error, and determine whether the amount of forecasting error corresponds to an anomalous event associated with the dataset of data instances. Methods and computer program products are also provided.

Abstract: The present application provides tricyclic urea compounds that modulate the activity of the V617F variant of JAK2, which are useful in the treatment of various diseases, including cancer.

Abstract: An aeronautical aluminum alloy minimum-quantity-lubrication milling machining device includes a machine tool worktable and spindle connected with a machine tool power system. The spindle is connected with a tool holder that is fixed with a cutting tool. The machine tool worktable is provided with a machine tool fixture, the tool holder is connected with a minimum-quantity-lubrication mechanism, the machine tool fixture includes a fixture body that is fixedly provided with a limit block for contact with two adjacent side surfaces of a workpiece, the fixture body is provided with a plurality of clamping elements capable of pressing the workpiece against an upper surface of the fixture body, and a top of the clamping element is provided with a detection member for detecting a relative position between the clamping element and the spindle. The device can avoid interference and contact between a nozzle and the clamping element.

Abstract: A liquid crystal display panel is provided. The liquid crystal display panel includes: a first substrate and a second substrate arranged opposite to each other, and a liquid crystal layer and a plurality of strip-shaped spacers disposed between the first substrate and the second substrate. In the liquid crystal display panel, there is an overlapping area between an orthographic projection of a first signal line on a target base and an orthographic projection of a second signal line on the target base, and an orthographic projection of the strip-shaped spacer on the target base is not overlapped with the overlapping area.

Abstract: A method includes bonding a first device die and a second device die to an interconnect die. The interconnect die includes a first portion over and bonded to the first device die, and a second portion over and bonded to the second device die. The interconnect die electrically connects the first device die to the second device die. The method further includes encapsulating the interconnect die in an encapsulating material, and forming a plurality of redistribution lines over the interconnect die.

Abstract: A layer stack including a first bonding dielectric material layer, a dielectric metal oxide layer, and a second bonding dielectric material layer is formed over a top surface of a substrate including a substrate semiconductor layer. A conductive material layer is formed by depositing a conductive material over the second bonding dielectric material layer. The substrate semiconductor layer is thinned by removing portions of the substrate semiconductor layer that are distal from the layer stack, whereby a remaining portion of the substrate semiconductor layer includes a top semiconductor layer. A semiconductor device may be formed on the top semiconductor layer.

Abstract: Provided are FinFET devices and methods of forming the same. A dummy gate having gate spacers on opposing sidewalls thereof is formed over a substrate. A dielectric layer is formed around the dummy gate. An upper portion of the dummy gate is removed and upper portions of the gate spacers are removed, so as to form a first opening in the dielectric layer. A lower portion of the dummy gate is removed to form a second opening below the first opening. A metal layer is formed in the first and second openings. The metal layer is partially removed to form a metal gate.

Abstract: Disclosed are a control information sending method and detecting method, a base station, a terminal, and a computer storage medium. The method includes: a base station determining first-type physical layer control information, which is used for indicating a first-type control parameter of a second-type physical layer control channel; determining second-type physical layer control information, which is used for indicating a second-type control parameter of a data channel; sending the first-type physical layer control information; and sending the second-type physical layer control information on the second-type physical layer control channel.

Abstract: An electronic device includes a substrate, a driving element, a first insulating layer, a pixel electrode layer, and a common electrode layer. The driving element is disposed on the substrate. The first insulating layer is disposed on the driving element. The pixel electrode layer is disposed on the first insulating layer. The first insulating layer comprises a hole, and the pixel electrode layer is electrically connected to the driving element through the hole. The common electrode layer is disposed on the pixel electrode layer. The common electrode layer comprises a slit, and the slit has an edge, and the edge is disposed in the hole.

Abstract: A method of contrastive learning comprises: determining, based on a model construction criterion, a first encoder for a first modality and a second encoder for a second modality; constructing a first contrastive learning model, the first contrastive learning model comprising the first encoder and a third encoder for the second modality, and a model capacity of the third encoder being greater than a model capacity of the second encoder; performing pre-training of the first contrastive learning model based on a first training dataset for the first modality and the second modality; and providing the pre-trained first encoder in the pre-trained first contrastive learning model for a downstream task. Because only the model capacity of one encoder is increased in the pre-training stage, model performance may be improved without increasing model training overhead during downstream task fine-tuning and model running overhead during model application.

Abstract: This application provides an exception locating method includes: A first device receives a first exception signal which changes abruptly from a first communication signal when the first communication signal passes through a location of an exception on a first line. The first device receives exception alarm information which is sent by the second device after the second device receives a second exception signal, the second exception signal is a signal that changes abruptly from a second communication signal when the second communication signal passes through the location of the exception. The first device determines the location of the exception based on a first time interval, where the first time interval is an interval between time at which the first exception signal is received and time at which the exception alarm information is received.