Patents by Inventor Minjia Zhang

Minjia Zhang has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Patent number: 11526728
    Abstract: Systems, methods, and computer-executable instructions for determining a computation schedule for a recurrent neural network (RNN). A matrix multiplication (MM) directed-acyclic graph (DAG) is received for the RNN. Valid phased computation schedules for the RNN are generated. Each of the valid phase computation schedule includes an ordering of MM operations. For each of the plurality of valid phased computation schedules, each of the MM operations is partitioned to processor cores based on L3 cache to L2 cache data movement. The RNN is executed based on the valid phased computation schedules. A final computation schedule is stored. The final computation schedule is used for future executions of the RNN.
    Type: Grant
    Filed: June 26, 2018
    Date of Patent: December 13, 2022
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Minjia Zhang, Samyam Rajbhandari, Wenhan Wang, Yuxiong He
  • Patent number: 11216459
    Abstract: A method for semantic search includes receiving a query vector including a semantic feature value for each of a plurality of semantic feature dimensions. A cluster is selected from a plurality of different candidate clusters held in a relatively fast memory, each candidate cluster including a plurality of compressed answer vectors. A subset of the plurality of compressed answer vectors are promoted as candidate answers. For each of the candidate answers, a corresponding uncompressed answer vector is retrieved from a relatively slower memory. A selected answer is promoted from among the candidate answers.
    Type: Grant
    Filed: March 25, 2019
    Date of Patent: January 4, 2022
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Minjia Zhang, Yuxiong He
  • Publication number: 20200311077
    Abstract: A method for semantic search includes receiving a query vector including a semantic feature value for each of a plurality of semantic feature dimensions. A cluster is selected from a plurality of different candidate clusters held in a relatively fast memory, each candidate cluster including a plurality of compressed answer vectors. A subset of the plurality of compressed answer vectors are promoted as candidate answers. For each of the candidate answers, a corresponding uncompressed answer vector is retrieved from a relatively slower memory. A selected answer is promoted from among the candidate answers.
    Type: Application
    Filed: March 25, 2019
    Publication date: October 1, 2020
    Applicant: Microsoft Technology Licensing, LLC
    Inventors: Minjia ZHANG, Yuxiong HE
  • Patent number: 10585988
    Abstract: Systems, methods, and computer-executable instructions for approximating a softmax layer are disclosed. A small world graph that includes a plurality of nodes is constructed for a vocabulary of a natural language model. A context vector is transformed. The small world graph is searched using the transformed context vector to identify a top-K hypothesis. A distance from the context vector for each of the top-K hypothesis is determined. The distance is transformed to an original inner product space. A softmax distribution is computed for the softmax layer over the inner product space of the top-K hypothesis. The softmax layer is useful for determining a next word in a speech recognition or machine translation.
    Type: Grant
    Filed: June 28, 2018
    Date of Patent: March 10, 2020
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Minjia Zhang, Xiaodong Liu, Wenhan Wang, Jianfeng Gao, Yuxiong He
  • Publication number: 20190377792
    Abstract: Systems, methods, and computer-executable instructions for approximating a softmax layer are disclosed. A small world graph that includes a plurality of nodes is constructed for a vocabulary of a natural language model. A context vector is transformed. The small world graph is searched using the transformed context vector to identify a top-K hypothesis. A distance from the context vector for each of the top-K hypothesis is determined. The distance is transformed to an original inner product space. A softmax distribution is computed for the softmax layer over the inner product space of the top-K hypothesis. The softmax layer is useful for determining a next word in a speech recognition or machine translation.
    Type: Application
    Filed: June 28, 2018
    Publication date: December 12, 2019
    Inventors: Minjia Zhang, Xiaodong Liu, Wenhan Wang, Jianfeng Gao, Yuxiong He
  • Publication number: 20190311245
    Abstract: Systems, methods, and computer-executable instructions for determining a computation schedule for a recurrent neural network (RNN). A matrix multiplication (MM) directed-acyclic graph (DAG) is received for the RNN. Valid phased computation schedules for the RNN are generated. Each of the valid phase computation schedule includes an ordering of MM operations. For each of the plurality of valid phased computation schedules, each of the MM operations is partitioned to processor cores based on L3 cache to L2 cache data movement. The RNN is executed based on the valid phased computation schedules. A final computation schedule is stored. The final computation schedule is used for future executions of the RNN.
    Type: Application
    Filed: June 26, 2018
    Publication date: October 10, 2019
    Inventors: Minjia Zhang, Samyam Rajbhandari, Wenhan Wang, Yuxiong He