Patents by Inventor David Benjamin Pfau

David Benjamin Pfau 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).

  • Antisymmetric neural networks
    Patent number: 12248861
    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing inputs using antisymmetric neural networks.
    Type: Grant
    Filed: September 3, 2020
    Date of Patent: March 11, 2025
    Assignee: DeepMind Technologies Limited
    Inventors: David Benjamin Pfau, James Spencer, Alexander Graeme de Garis Matthews
  • TRAINING MACHINE LEARNING MODELS BY DETERMINING UPDATE RULES USING NEURAL NETWORKS
    Publication number: 20230376771
    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media for training machine learning models. One method includes obtaining a machine learning model, wherein the machine learning model comprises one or more model parameters, and the machine learning model is trained using gradient descent techniques to optimize an objective function; determining an update rule for the model parameters using a recurrent neural network (RNN); and applying a determined update rule for a final time step in a sequence of multiple time steps to the model parameters.
    Type: Application
    Filed: March 8, 2023
    Publication date: November 23, 2023
    Inventors: Misha Man Ray Denil, Tom Schaul, Marcin Andrychowicz, Joao Ferdinando Gomes de Freitas, Sergio Gomez Colmenarejo, Matthew William Hoffman, David Benjamin Pfau
  • Training machine learning models by determining update rules using recurrent neural networks
    Patent number: 11615310
    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media for training machine learning models. One method includes obtaining a machine learning model, wherein the machine learning model comprises one or more model parameters, and the machine learning model is trained using gradient descent techniques to optimize an objective function; determining an update rule for the model parameters using a recurrent neural network (RNN); and applying a determined update rule for a final time step in a sequence of multiple time steps to the model parameters.
    Type: Grant
    Filed: May 19, 2017
    Date of Patent: March 28, 2023
    Assignee: DeepMind Technologies Limited
    Inventors: Misha Man Ray Denil, Tom Schaul, Marcin Andrychowicz, Joao Ferdinando Gomes de Freitas, Sergio Gomez Colmenarejo, Matthew William Hoffman, David Benjamin Pfau
  • TRAINING SPECTRAL INFERENCE NEURAL NETWORKS USING BILEVEL OPTIMIZATION
    Publication number: 20210166131
    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training a neural network having a plurality of network parameters and being configured to process an input data item to generate a feature representation comprising a values for each of a plurality of features of the input data item.
    Type: Application
    Filed: June 6, 2019
    Publication date: June 3, 2021
    Inventors: David Benjamin Pfau, Stig Petersen, Ashish Agarwal, David Barrett, Kimberly Stachenfeld
  • ANTISYMMETRIC NEURAL NETWORKS
    Publication number: 20210064961
    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for processing inputs using antisymmetric neural networks.
    Type: Application
    Filed: September 3, 2020
    Publication date: March 4, 2021
    Inventors: David Benjamin Pfau, James Spencer, Alexander Graeme de Garis Matthews
  • TRAINING MACHINE LEARNING MODELS
    Publication number: 20190220748
    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media for training machine learning models. One method includes obtaining a machine learning model, wherein the machine learning model comprises one or more model parameters, and the machine learning model is trained using gradient descent techniques to optimize an objective function; determining an update rule for the model parameters using a recurrent neural network (RNN); and applying a determined update rule for a final time step in a sequence of multiple time steps to the model parameters.
    Type: Application
    Filed: May 19, 2017
    Publication date: July 18, 2019
    Inventors: Misha Man Ray Denil, Tom Schaul, Marcin Andrychowicz, Joao Ferdinando Gomes de Freitas, Sergio Gomez Colmenarejo, Matthew William Hoffman, David Benjamin Pfau