Abstract: Described herein are embodiments of a reinforcement learning based large-scale multi-objective ranking system. Embodiments of the system may be used for optimizing short-video recommendation on a video sharing platform. Multiple competing ranking objective and implicit selection bias in user feedback are the main challenges in real-world platform. In order to address those challenges, multi-gate mixture of experts (MMoE) and soft actor critic (SAC) are integrated together into a MMoE_SAC system. Experiment results demonstrate that embodiments of the MMoE_SAC system may greatly reduce a loss function compared to systems only based on single strategies.

Abstract: Described herein are embodiments of a framework named as total correlation variational autoencoder (TC_VAE) to disentangle syntax and semantics by making use of total correlation penalties of KL divergences. One or more Kullback-Leibler (KL) divergence terms in a loss for a variational autoencoder are discomposed so that generated hidden variables may be separated. Embodiments of the TC_VAE framework were examined on semantic similarity tasks and syntactic similarity tasks. Experimental results show that better disentanglement between syntactic and semantic representations have been achieved compared with state-of-the-art (SOTA) results on the same data sets in similar settings.

Abstract: Embodiments of three-dimensional (3D) memory devices with a 3D NAND memory array having a plurality of pages and an on-die data buffer coupled to the memory array on a same chip and configured to buffer a plurality of batches of program data between a host and the memory array. The on-die data buffer may include SRAM cells. The 3D memory device also includes a controller coupled to the on-die data buffer on the same chip. The controller may be configured to receive control instructions for performing a first pass program and a second pass program on memory cells in a page. The controller may also be configured to buffer, in the on-die data buffer, first program data for a first pass program and second program data for a second pass program from a host and retrieve the first program data from the on-die data buffer.

Abstract: Provided is use of Penicillium sclerotiorum SCAUMCX01, under accession number GDMCC No. 60249, in the preparation of plant growth regulators or inducers. The plant growth regulators or inducers containing Penicillium sclerotiorum, Penicillium sclerotiorum spores, a culture solution thereof and an extract of the culture solution have the effects of promoting the early germination of seeds, promoting the growth of the underground roots of crops and the branching of above-ground parts of crops, improving the bioavailability of crops, delaying the aging progress of crops, and enhancing the growth of crops.

Abstract: Aspects of the present disclose provide various methods and apparatuses for communicating, controlling, and configuring component carrier and bandwidth part (BWP). A scheduling entity receives a capability report from a user equipment (UE). The capability report indicates a capability of the UE to utilize at least one of carrier aggregation (CA) or one or more bandwidth parts. The scheduling entity transmits a command to the UE to reconfigure at least one of a CA configuration or a bandwidth part (BWP) configuration. The scheduling entity determines an anticipated response timing of an acknowledgment (ACK) of the command based on the capability report received from the UE. The scheduling entity receives the ACK according to the anticipated response timing.

Abstract: An amplifier circuit structure can include an amplifier located in a main path, and a first switch located in a bypass. One end of a second switch is a signal output end of the amplifier circuit structure, and the other end of the second switch is configured to selectively connect to a signal output end of the bypass or a signal output end of the main path. The first and second switches are configured to control their respective operating states when a first instruction is received, such that the main path is connected to the signal input end and the signal output end of the amplifier circuit structure; and to control their respective operating states when a second instruction is received, such that the bypass is connected to the signal input end of the amplifier circuit structure and the signal output end of the amplifier circuit structure.

Abstract: Techniques are described herein related to communicating a quality of service (QoS) class identifier (QCI) for configuring physical (PHY) layer signaling. The QCI may be used to configure one or more transmission parameters of a device (e.g., a user equipment (UE) or a base station) at a PHY layer. The PHY layer may select the one or more transmission parameters to meet the QoS conditions of a given QoS class of a transport block being communicated. A modulation and coding scheme (MCS) value may be used as the QCI for the PHY layer, a channel quality indicator (CQI) value may be used as the QCI for the PHY layer, or a cell radio network temporary identifier (C-RNTI) may be used as the QCI for the PHY layer. In some cases, the QCI may be a single bit included in a downlink control message.

Abstract: Methods, systems, and devices for wireless communications are described that provide for service indication via downlink control information (DCI) in which different services have a same DCI size. Techniques provide also that a DCI format may be indicated by a masking operation on a cyclic redundancy check (CRC) of the DCI. The DCI may include an indicator field that may indicate the service of the associated operation. The indication may also be provided via different monitoring occasions of the DCI transmissions that are associated with different services. Different DCI search spaces also may be associated with different services and provide the service indication. Additionally or alternatively, different scrambling sequences used on all or a portion of control information or demodulation reference signal transmissions may be associated with different services.

Abstract: Systems and techniques are disclosed for selecting service-specific cells that considers additional information about service availability at the cells. A base station stores service availability information in system information. A UE searching for a service-specific cell on which to camp receives the system information with service availability. The UE analyzes its service requirements against the service availability as well as measured radio conditions and selects a cell that enables service-specific support on which to camp. When a change occurs to availability of the service support at the base station, it notifies the camped UE of the change and the UE obtains the changed system information to determine whether to reselect a different cell on which to camp. When a service-specific cell is not available, the UE may select a suitable cell for normal service and periodically perform cell reselection in an attempt to again camp on a service-specific cell.

Abstract: A crusher includes a blade assembly and a drive assembly. The blade assembly includes a first blade and a second blade. The drive assembly is respectively connected with the first blade and the second blade, and is configured to drive the first blade and the second blade to rotate. The first blade is configured to disperse and perform primary cutting of the materials. The second blade is configured to perform secondary cutting of the materials. The first blade makes the materials fully disperse, so that the second blade can fully cut the materials. In addition, the first blade initially cuts the materials, so that the formed fragments can achieve a better fertilization effect.

Abstract: Methods, systems, and devices for wireless communications are described. An uplink preemption indication (PI) may be transmitted from a serving cell to a user equipment (UE), where the uplink PI indicates that uplink traffic is to be preempted. In some cases, the UE may communicate with the serving cell on a first component carrier (CC) according to a first time division duplexing (TDD) configuration, where the first TDD configuration includes uplink and downlink portions. The serving cell may transmit the uplink PI on a second CC, and the UE may monitor for and receive the uplink PI during downlink portions of the second CC during uplink portions of the first CC. In some cases, the UE may monitor the second CC in accordance with a frequency division duplexing (FDD) configuration. Additionally or alternatively, the UE may monitor the second CC in accordance with a second complementary TDD configuration.

Abstract: Presented herein are embodiments of an unsupervised cross-lingual sentiment classification model (which may be referred to as multi-view encoder-classifier (MVEC)) that leverages an unsupervised machine translation (UMT) system and a language discriminator. Unlike previous language model (LM)-based fine-tuning approaches that adjust parameters solely based on the classification error on training data, embodiments employ an encoder-decoder framework of an UMT as a regularization component on the shared network parameters. In one or more embodiments, the cross-lingual encoder of embodiments learns a shared representation, which is effective for both reconstructing input sentences of two languages and generating more representative views from the input for classification. Experiments on five language pairs verify that an MVEC embodiment significantly outperforms other models for 8/11 sentiment classification tasks.

Abstract: Concept graphs are created as universal taxonomies for text understanding in the open domain knowledge. The nodes in concept graphs include both entities and concepts. The edges are from entities to concepts, showing that an entity is an instance of a concept. Presented herein are embodiments that handle the task of learning interpretable relationships from open domain facts to enrich and refine concept graphs. In one or more embodiments, the Bayesian network structures are learned from open domain facts as the interpretable relationships between relations of facts and concepts of entities. Extensive experiments were conducted on English and Chinese datasets. Compared to the state-of-the-art methods, the learned network structures improve the identification of concepts for entities based on the relations of entities on both English and Chinese datasets.

Abstract: Embodiments of 3D memory devices with a static random-access memory (SRAM) and fabrication methods thereof are disclosed herein. In one example, the method for operating a 3D memory device having an input/output circuit, an array of SRAM cells, and an array of 3D NAND memory strings in a same chip. The method may include transferring data through the input/output circuit to the array of SRAM cells, storing the data in the array of SRAM cells, and programming the data into the array of 3D NAND memory strings from the array of SRAM cells.

Abstract: Methods, systems, and devices for wireless communications are described. Generally, the described techniques provide for identification of a slot format for a set of slots where at least one slot format of the set is an subband full-duplex slot (e.g., includes uplink and downlink resources separated in a frequency domain). The base station may transmit one or more semi-persistent scheduling configurations to a UE. The UE may identify semi-persistent scheduling occasions for receiving a downlink communication based at least in part on the SPS configurations matching the format of the slot, the periodicity of the slot formats of the set, or other considerations. The UE and the base station may communicate based at least in part on the determined semi-persistent scheduling occasions.

Abstract: The present disclosure provides a bulk acoustic wave resonator, a manufacturing method thereof, and a filter, wherein the bulk acoustic wave resonator includes: a substrate; an acoustic reflection unit on the substrate; a piezoelectric stack structure on the acoustic reflection unit; and a pad on the piezoelectric stack structure; wherein the pad has an overlapping region with the acoustic reflection unit. The acoustic wave resonator, the manufacturing method thereof and the filter of the present disclosure can effectively reduce connection resistance of the bulk acoustic wave resonator, thereby reducing insertion loss of the filter.

Abstract: An impedance adjustment circuit is connected in parallel with a bias current output end of a bias circuit. The bias circuit is configured to provide bias current to a first circuit unit. The impedance adjustment circuit is configured to adjust source impedance of the first circuit unit.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment may determine a block error rate (BLER) target for communications associated with the user equipment; determine a resource allocation pattern for transmission of channel state information reference signals (CSI-RS) based at least in part on the BLER target; and monitor one or more resources, indicated by the resource allocation pattern, for the CSI-RS. Numerous other aspects are provided.

Abstract: In an aspect of the disclosure, a method, a computer-readable medium, and an apparatus are provided. In certain configurations, the apparatus may include an eMBB UE. The apparatus may monitor a single preconfigured mini-slot in each of a plurality of slots for a PI. The apparatus may receive the PI in the single preconfigured mini-slot of a first slot of the plurality of slots. In certain aspects, the PI may include a resource index associated with a reduced transmission power by the UE. In certain other aspects, the resource index may include one or more resources in a second slot. The apparatus may transmit at least one packet with the reduced transmission power on the one or more resources in the second slot or refraining from transmitting the at least one packet on the one or more resources in the second slot.

Abstract: Methods related to wireless communication systems and transmission and/or reception of synchronization signal block (SSB) transmissions (e.g., in full-duplex slots) are provided. For example, a wireless communication device receives a synchronization signal block (SSB) transmission schedule. The SSB transmission schedule is at least partially overlapping with an uplink (UL) frequency portion of a full-duplex slot. The wireless communication device receives a first portion of a SSB in a downlink (DL) frequency portion of the full-duplex slot; and receiving a second portion of the SSB in the UL frequency portion of the full-duplex slot based on the SSB being scheduled in the full-duplex slot and at least partially overlapping with the UL frequency portion of the full-duplex slot. Other features are also claimed and described.