Patents by Inventor Daniel Pieter Wierstra
Daniel Pieter Wierstra 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).
-
Publication number: 20240046106Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using multi-task neural networks. One of the methods includes receiving a first network input and data identifying a first machine learning task to be performed on the first network input; selecting a path through the plurality of layers in a super neural network that is specific to the first machine learning task, the path specifying, for each of the layers, a proper subset of the modular neural networks in the layer that are designated as active when performing the first machine learning task; and causing the super neural network to process the first network input using (i) for each layer, the modular neural networks in the layer that are designated as active by the selected path and (ii) the set of one or more output layers corresponding to the identified first machine learning task.Type: ApplicationFiled: October 16, 2023Publication date: February 8, 2024Inventors: Daniel Pieter Wierstra, Chrisantha Thomas Fernando, Alexander Pritzel, Dylan Sunil Banarse, Charles Blundell, Andrei-Alexandru Rusu, Yori Zwols, David Ha
-
Patent number: 11803750Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training an actor neural network used to select actions to be performed by an agent interacting with an environment. One of the methods includes obtaining a minibatch of experience tuples; and updating current values of the parameters of the actor neural network, comprising: for each experience tuple in the minibatch: processing the training observation and the training action in the experience tuple using a critic neural network to determine a neural network output for the experience tuple, and determining a target neural network output for the experience tuple; updating current values of the parameters of the critic neural network using errors between the target neural network outputs and the neural network outputs; and updating the current values of the parameters of the actor neural network using the critic neural network.Type: GrantFiled: September 14, 2020Date of Patent: October 31, 2023Assignee: DeepMind Technologies LimitedInventors: Timothy Paul Lillicrap, Jonathan James Hunt, Alexander Pritzel, Nicolas Manfred Otto Heess, Tom Erez, Yuval Tassa, David Silver, Daniel Pieter Wierstra
-
Patent number: 11790238Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using multi-task neural networks. One of the methods includes receiving a first network input and data identifying a first machine learning task to be performed on the first network input; selecting a path through the plurality of layers in a super neural network that is specific to the first machine learning task, the path specifying, for each of the layers, a proper subset of the modular neural networks in the layer that are designated as active when performing the first machine learning task; and causing the super neural network to process the first network input using (i) for each layer, the modular neural networks in the layer that are designated as active by the selected path and (ii) the set of one or more output layers corresponding to the identified first machine learning task.Type: GrantFiled: August 17, 2020Date of Patent: October 17, 2023Assignee: DeepMind Technologies LimitedInventors: Daniel Pieter Wierstra, Chrisantha Thomas Fernando, Alexander Pritzel, Dylan Sunil Banarse, Charles Blundell, Andrei-Alexandru Rusu, Yori Zwols, David Ha
-
Patent number: 11336908Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for compressing images using neural networks. One of the methods includes receiving an image; processing the image using an encoder neural network, wherein the encoder neural network is configured to receive the image and to process the image to generate an output defining values of a first number of latent variables that each represent a feature of the image; generating a compressed representation of the image using the output defining the values of the first number of latent variables; and providing the compressed representation of the image for use in generating a reconstruction of the image.Type: GrantFiled: September 27, 2019Date of Patent: May 17, 2022Assignee: DeepMind Technologies LimitedInventors: Daniel Pieter Wierstra, Karol Gregor, Frederic Olivier Besse
-
Patent number: 11328183Abstract: A neural network system is proposed. The neural network can be trained by model-based reinforcement learning to select actions to be performed by an agent interacting with an environment, to perform a task in an attempt to achieve a specified result. The system may comprise at least one imagination core which receives a current observation characterizing a current state of the environment, and optionally historical observations, and which includes a model of the environment. The imagination core may be configured to output trajectory data in response to the current observation, and/or historical observations. The trajectory data comprising a sequence of future features of the environment imagined by the imagination core. The system may also include a rollout encoder to encode the features, and an output stage to receive data derived from the rollout embedding and to output action policy data for identifying an action based on the current observation.Type: GrantFiled: September 14, 2020Date of Patent: May 10, 2022Assignee: DeepMind Technologies LimitedInventors: Daniel Pieter Wierstra, Yujia Li, Razvan Pascanu, Peter William Battaglia, Theophane Guillaume Weber, Lars Buesing, David Paul Reichert, Arthur Clement Guez, Danilo Jimenez Rezende, Adrià Puigdomènech Badia, Oriol Vinyals, Nicolas Manfred Otto Heess, Sebastien Henri Andre Racaniere
-
Patent number: 11200482Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for environment simulation. In one aspect, a system comprises a recurrent neural network configured to, at each of a plurality of time steps, receive a preceding action for a preceding time step, update a preceding initial hidden state of the recurrent neural network from the preceding time step using the preceding action, update a preceding cell state of the recurrent neural network from the preceding time step using at least the initial hidden state for the time step, and determine a final hidden state for the time step using the cell state for the time step. The system further comprises a decoder neural network configured to receive the final hidden state for the time step and process the final hidden state to generate a predicted observation characterizing a predicted state of the environment at the time step.Type: GrantFiled: June 5, 2020Date of Patent: December 14, 2021Assignee: DeepMind Technologies LimitedInventors: Daniel Pieter Wierstra, Shakir Mohamed, Silvia Chiappa, Sebastien Henri Andre Racaniere
-
Publication number: 20210089834Abstract: A neural network system is proposed to select actions to be performed by an agent interacting with an environment to perform a task in an attempt to achieve a specified result. The system may include a controller to receive state data and context data, and to output action data. The system may also include an imagination module to receive the state and action data, and to output consequent state data. The system may also include a manager to receive the state data and the context data, and to output route data which defines whether the system is to execute an action or to imagine. The system may also include a memory to store the context data.Type: ApplicationFiled: December 7, 2020Publication date: March 25, 2021Inventors: Daniel Pieter Wierstra, Yujia Li, Razvan Pascanu, Peter William Battaglia, Theophane Guillaume Weber, Lars Buesing, David Paul Reichert, Oriol Vinyals, Nicolas Manfred Otto Heess, Sebastien Henri Andre Racaniere
-
Publication number: 20210073594Abstract: A neural network system is proposed. The neural network can be trained by model-based reinforcement learning to select actions to be performed by an agent interacting with an environment, to perform a task in an attempt to achieve a specified result. The system may comprise at least one imagination core which receives a current observation characterizing a current state of the environment, and optionally historical observations, and which includes a model of the environment. The imagination core may be configured to output trajectory data in response to the current observation, and/or historical observations. The trajectory data comprising a sequence of future features of the environment imagined by the imagination core. The system may also include a rollout encoder to encode the features, and an output stage to receive data derived from the rollout embedding and to output action policy data for identifying an action based on the current observation.Type: ApplicationFiled: September 14, 2020Publication date: March 11, 2021Inventors: Daniel Pieter Wierstra, Yujia Li, Razvan Pascanu, Peter William Battaglia, Theophane Guillaume Weber, Lars Buesing, David Paul Reichert, Arthur Clement Guez, Danilo Jimenez Rezende, Adrià Puigdomènech Badia, Oriol Vinyals, Nicolas Manfred Otto Heess, Sebastien Henri Andre Racaniere
-
Publication number: 20200410351Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training an actor neural network used to select actions to be performed by an agent interacting with an environment. One of the methods includes obtaining a minibatch of experience tuples; and updating current values of the parameters of the actor neural network, comprising: for each experience tuple in the minibatch: processing the training observation and the training action in the experience tuple using a critic neural network to determine a neural network output for the experience tuple, and determining a target neural network output for the experience tuple; updating current values of the parameters of the critic neural network using errors between the target neural network outputs and the neural network outputs; and updating the current values of the parameters of the actor neural network using the critic neural network.Type: ApplicationFiled: September 14, 2020Publication date: December 31, 2020Inventors: Timothy Paul Lillicrap, Jonathan James Hunt, Alexander Pritzel, Nicolas Manfred Otto Heess, Tom Erez, Yuval Tassa, David Silver, Daniel Pieter Wierstra
-
Patent number: 10860895Abstract: A neural network system is proposed to select actions to be performed by an agent interacting with an environment to perform a task in an attempt to achieve a specified result. The system may include a controller to receive state data and context data, and to output action data. The system may also include an imagination module to receive the state and action data, and to output consequent state data. The system may also include a manager to receive the state data and the context data, and to output route data which defines whether the system is to execute an action or to imagine. The system may also include a memory to store the context data.Type: GrantFiled: November 19, 2019Date of Patent: December 8, 2020Assignee: DeepMind Technologies LimitedInventors: Daniel Pieter Wierstra, Yujia Li, Razvan Pascanu, Peter William Battaglia, Theophane Guillaume Weber, Lars Buesing, David Paul Reichert, Oriol Vinyals, Nicolas Manfred Otto Heess, Sebastien Henri Andre Racaniere
-
Publication number: 20200380372Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using multi-task neural networks. One of the methods includes receiving a first network input and data identifying a first machine learning task to be performed on the first network input; selecting a path through the plurality of layers in a super neural network that is specific to the first machine learning task, the path specifying, for each of the layers, a proper subset of the modular neural networks in the layer that are designated as active when performing the first machine learning task; and causing the super neural network to process the first network input using (i) for each layer, the modular neural networks in the layer that are designated as active by the selected path and (ii) the set of one or more output layers corresponding to the identified first machine learning task.Type: ApplicationFiled: August 17, 2020Publication date: December 3, 2020Inventors: Daniel Pieter Wierstra, Chrisantha Thomas Fernando, Alexander Pritzel, Dylan Sunil Banarse, Charles Blundell, Andrei-Alexandru Rusu, Yori Zwols, David Ha
-
Publication number: 20200342289Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for environment simulation. In one aspect, a system comprises a recurrent neural network configured to, at each of a plurality of time steps, receive a preceding action for a preceding time step, update a preceding initial hidden state of the recurrent neural network from the preceding time step using the preceding action, update a preceding cell state of the recurrent neural network from the preceding time step using at least the initial hidden state for the time step, and determine a final hidden state for the time step using the cell state for the time step. The system further comprises a decoder neural network configured to receive the final hidden state for the time step and process the final hidden state to generate a predicted observation characterizing a predicted state of the environment at the time step.Type: ApplicationFiled: June 5, 2020Publication date: October 29, 2020Inventors: Daniel Pieter Wierstra, Shakir Mohamed, Silvia Chiappa, Sebastien Henri Andre Racaniere
-
Patent number: 10776692Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training an actor neural network used to select actions to be performed by an agent interacting with an environment. One of the methods includes obtaining a minibatch of experience tuples; and updating current values of the parameters of the actor neural network, comprising: for each experience tuple in the minibatch: processing the training observation and the training action in the experience tuple using a critic neural network to determine a neural network output for the experience tuple, and determining a target neural network output for the experience tuple; updating current values of the parameters of the critic neural network using errors between the target neural network outputs and the neural network outputs; and updating the current values of the parameters of the actor neural network using the critic neural network.Type: GrantFiled: July 22, 2016Date of Patent: September 15, 2020Assignee: DeepMind Technologies LimitedInventors: Timothy Paul Lillicrap, Jonathan James Hunt, Alexander Pritzel, Nicolas Manfred Otto Heess, Tom Erez, Yuval Tassa, David Silver, Daniel Pieter Wierstra
-
Patent number: 10776670Abstract: A neural network system is proposed. The neural network can be trained by model-based reinforcement learning to select actions to be performed by an agent interacting with an environment, to perform a task in an attempt to achieve a specified result. The system may comprise at least one imagination core which receives a current observation characterizing a current state of the environment, and optionally historical observations, and which includes a model of the environment. The imagination core may be configured to output trajectory data in response to the current observation, and/or historical observations. The trajectory data comprising a sequence of future features of the environment imagined by the imagination core. The system may also include a rollout encoder to encode the features, and an output stage to receive data derived from the rollout embedding and to output action policy data for identifying an action based on the current observation.Type: GrantFiled: November 19, 2019Date of Patent: September 15, 2020Assignee: DeepMind Technologies LimitedInventors: Daniel Pieter Wierstra, Yujia Li, Razvan Pascanu, Peter William Battaglia, Theophane Guillaume Weber, Lars Buesing, David Paul Reichert, Arthur Clement Guez, Danilo Jimenez Rezende, Adrià Puigdomènech Badia, Oriol Vinyals, Nicolas Manfred Otto Heess, Sebastien Henri Andre Racaniere
-
Patent number: 10748065Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using multi-task neural networks. One of the methods includes receiving a first network input and data identifying a first machine learning task to be performed on the first network input; selecting a path through the plurality of layers in a super neural network that is specific to the first machine learning task, the path specifying, for each of the layers, a proper subset of the modular neural networks in the layer that are designated as active when performing the first machine learning task; and causing the super neural network to process the first network input using (i) for each layer, the modular neural networks in the layer that are designated as active by the selected path and (ii) the set of one or more output layers corresponding to the identified first machine learning task.Type: GrantFiled: July 30, 2019Date of Patent: August 18, 2020Assignee: DeepMind Technologies LimitedInventors: Daniel Pieter Wierstra, Chrisantha Thomas Fernando, Alexander Pritzel, Dylan Sunil Banarse, Charles Blundell, Andrei-Alexandru Rusu, Yori Zwols, David Ha
-
Patent number: 10713559Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for environment simulation. In one aspect, a system comprises a recurrent neural network configured to, at each of a plurality of time steps, receive a preceding action for a preceding time step, update a preceding initial hidden state of the recurrent neural network from the preceding time step using the preceding action, update a preceding cell state of the recurrent neural network from the preceding time step using at least the initial hidden state for the time step, and determine a final hidden state for the time step using the cell state for the time step. The system further comprises a decoder neural network configured to receive the final hidden state for the time step and process the final hidden state to generate a predicted observation characterizing a predicted state of the environment at the time step.Type: GrantFiled: May 3, 2019Date of Patent: July 14, 2020Assignee: DeepMind Technologies LimitedInventors: Daniel Pieter Wierstra, Shakir Mohamed, Silvia Chiappa, Sebastien Henri Andre Racaniere
-
Publication number: 20200090006Abstract: A neural network system is proposed. The neural network can be trained by model-based reinforcement learning to select actions to be performed by an agent interacting with an environment, to perform a task in an attempt to achieve a specified result. The system may comprise at least one imagination core which receives a current observation characterizing a current state of the environment, and optionally historical observations, and which includes a model of the environment. The imagination core may be configured to output trajectory data in response to the current observation, and/or historical observations. The trajectory data comprising a sequence of future features of the environment imagined by the imagination core. The system may also include a rollout encoder to encode the features, and an output stage to receive data derived from the rollout embedding and to output action policy data for identifying an action based on the current observation.Type: ApplicationFiled: November 19, 2019Publication date: March 19, 2020Inventors: Daniel Pieter Wierstra, Yujia Li, Razvan Pascanu, Peter William Battaglia, Theophane Guillaume Weber, Lars Buesing, David Paul Reichert, Arthur Clement Guez, Danilo Jimenez Rezende, Adrià Puigdomènech Badia, Oriol Vinyals, Nicolas Manfred Otto Heess, Sebastien Henri Andre Racaniere
-
Publication number: 20200082227Abstract: A neural network system is proposed to select actions to be performed by an agent interacting with an environment to perform a task in an attempt to achieve a specified result. The system may include a controller to receive state data and context data, and to output action data. The system may also include an imagination module to receive the state and action data, and to output consequent state data. The system may also include a manager to receive the state data and the context data, and to output route data which defines whether the system is to execute an action or to imagine. The system may also include a memory to store the context data.Type: ApplicationFiled: November 19, 2019Publication date: March 12, 2020Inventors: Daniel Pieter Wierstra, Yujia Li, Razvan Pascanu, Peter William Battaglia, Theophane Guillaume Weber, Lars Buesing, David Paul Reichert, Oriol Vinyals, Nicolas Manfred Otto Heess, Sebastien Henri Andre Racaniere
-
Publication number: 20200029084Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for compressing images using neural networks. One of the methods includes receiving an image; processing the image using an encoder neural network, wherein the encoder neural network is configured to receive the image and to process the image to generate an output defining values of a first number of latent variables that each represent a feature of the image; generating a compressed representation of the image using the output defining the values of the first number of latent variables; and providing the compressed representation of the image for use in generating a reconstruction of the image.Type: ApplicationFiled: September 27, 2019Publication date: January 23, 2020Inventors: Daniel Pieter Wierstra, Karol Gregor, Frederic Olivier Besse
-
Publication number: 20190354868Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using multi-task neural networks. One of the methods includes receiving a first network input and data identifying a first machine learning task to be performed on the first network input; selecting a path through the plurality of layers in a super neural network that is specific to the first machine learning task, the path specifying, for each of the layers, a proper subset of the modular neural networks in the layer that are designated as active when performing the first machine learning task; and causing the super neural network to process the first network input using (i) for each layer, the modular neural networks in the layer that are designated as active by the selected path and (ii) the set of one or more output layers corresponding to the identified first machine learning task.Type: ApplicationFiled: July 30, 2019Publication date: November 21, 2019Inventors: Daniel Pieter Wierstra, Chrisantha Thomas Fernando, Alexander Pritzel, Dylan Sunil Banarse, Charles Blundell, Andrei-Alexandru Rusu, Yori Zwols, David Ha