Patents by Inventor Mohammad Gheshlaghi Azar
Mohammad Gheshlaghi Azar 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: 11977983Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting an action to be performed by a reinforcement learning agent. The method includes obtaining an observation characterizing a current state of an environment. For each layer parameter of each noisy layer of a neural network, a respective noise value is determined. For each layer parameter of each noisy layer, a noisy current value for the layer parameter is determined from a current value of the layer parameter, a current value of a corresponding noise parameter, and the noise value. A network input including the observation is processed using the neural network in accordance with the noisy current values to generate a network output for the network input. An action is selected from a set of possible actions to be performed by the agent in response to the observation using the network output.Type: GrantFiled: September 14, 2020Date of Patent: May 7, 2024Assignee: DeepMind Technologies LimitedInventors: Mohammad Gheshlaghi Azar, Meire Fortunato, Bilal Piot, Olivier Claude Pietquin, Jacob Lee Menick, Volodymyr Mnih, Charles Blundell, Remi Munos
-
Publication number: 20230083486Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training an environment representation neural network of a reinforcement learning system controls an agent to perform a given task. In one aspect, the method includes: receiving a current observation input and a future observation input; generating, from the future observation input, a future latent representation of the future state of the environment; processing, using the environment representation neural network, to generate a current internal representation of the current state of the environment; generating, from the current internal representation, a predicted future latent representation; evaluating an objective function measuring a difference between the future latent representation and the predicted future latent representation; and determining, based on a determined gradient of the objective function, an update to the current values of the environment representation parameters.Type: ApplicationFiled: February 8, 2021Publication date: March 16, 2023Inventors: Zhaohan Guo, Mohammad Gheshlaghi Azar, Bernardo Avila Pires, Florent Altché, Jean-Bastien François Laurent Grill, Bilal Piot, Remi Munos
-
Patent number: 11604997Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a policy neural network. The policy neural network is used to select actions to be performed by an agent that interacts with an environment by receiving an observation characterizing a state of the environment and performing an action from a set of actions in response to the received observation. A trajectory is obtained from a replay memory, and a final update to current values of the policy network parameters is determined for each training observation in the trajectory. The final updates to the current values of the policy network parameters are determined from selected action updates and leave-one-out updates.Type: GrantFiled: June 11, 2018Date of Patent: March 14, 2023Assignee: DeepMind Technologies LimitedInventors: Marc Gendron-Bellemare, Mohammad Gheshlaghi Azar, Audrunas Gruslys, Remi Munos
-
Publication number: 20220092456Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for controlling an agent. One of the methods includes, while training a neural network used to control the agent, generating a reward value for the training as a measure of the divergence between the likelihood of the further observation under first and second statistical models of the environment, the first statistical model and second model being based on respective first and second histories of past observations and actions, the most recent observation in the first history being more recent than the most recent observation in the second history.Type: ApplicationFiled: January 23, 2020Publication date: March 24, 2022Inventors: BILAL PIOT, BERNARDO AVILA PIRES, MOHAMMAD GHESHLAGHI AZAR
-
Publication number: 20210383225Abstract: A computer-implemented method of training a neural network. The method comprises processing a first transformed view of a training data item, e.g. an image, with a target neural network to generate a target output, processing a second transformed view of the training data item, e.g. image, with an online neural network to generate a prediction of the target output, updating parameters of the online neural network to minimize an error between the prediction of the target output and the target output, and updating parameters of the target neural network based on the parameters of the online neural network. The method can effectively train an encoder neural network without using labelled training data items, and without using a contrastive loss, i.e. without needing “negative examples” which comprise transformed views of different data items.Type: ApplicationFiled: June 4, 2021Publication date: December 9, 2021Inventors: Jean-Bastien François Laurent Grill, Florian Strub, Florent Altché, Corentin Tallec, Pierre Richemond, Bernardo Avila Pires, Zhaohan Guo, Mohammad Gheshlaghi Azar, Bilal Piot, Remi Munos, Michal Valko
-
Publication number: 20210110271Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a policy neural network. The policy neural network is used to select actions to be performed by an agent that interacts with an environment by receiving an observation characterizing a state of the environment and performing an action from a set of actions in response to the received observation. A trajectory is obtained from a replay memory, and a final update to current values of the policy network parameters is determined for each training observation in the trajectory. The final updates to the current values of the policy network parameters are determined from selected action updates and leave-one-out updates.Type: ApplicationFiled: June 11, 2018Publication date: April 15, 2021Inventors: Marc Gendron-Bellemare, Mohammad Gheshlaghi Azar, Audrunas Gruslys, Remi Munos
-
Publication number: 20210065012Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting an action to be performed by a reinforcement learning agent. The method includes obtaining an observation characterizing a current state of an environment. For each layer parameter of each noisy layer of a neural network, a respective noise value is determined. For each layer parameter of each noisy layer, a noisy current value for the layer parameter is determined from a current value of the layer parameter, a current value of a corresponding noise parameter, and the noise value. A network input including the observation is processed using the neural network in accordance with the noisy current values to generate a network output for the network input. An action is selected from a set of possible actions to be performed by the agent in response to the observation using the network output.Type: ApplicationFiled: September 14, 2020Publication date: March 4, 2021Inventors: Mohammad Gheshlaghi Azar, Meire Fortunato, Bilal Piot, Olivier Claude Pietquin, Jacob Lee Menick, Volodymyr Mnih, Charles Blundell, Remi Munos
-
Patent number: 10839293Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting an action to be performed by a reinforcement learning agent. The method includes obtaining an observation characterizing a current state of an environment. For each layer parameter of each noisy layer of a neural network, a respective noise value is determined. For each layer parameter of each noisy layer, a noisy current value for the layer parameter is determined from a current value of the layer parameter, a current value of a corresponding noise parameter, and the noise value. A network input including the observation is processed using the neural network in accordance with the noisy current values to generate a network output for the network input. An action is selected from a set of possible actions to be performed by the agent in response to the observation using the network output.Type: GrantFiled: June 12, 2019Date of Patent: November 17, 2020Assignee: DeepMind Technologies LimitedInventors: Mohammad Gheshlaghi Azar, Meire Fortunato, Bilal Piot, Olivier Claude Pietquin, Jacob Lee Menick, Volodymyr Mnih, Charles Blundell, Remi Munos
-
Publication number: 20190362238Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting an action to be performed by a reinforcement learning agent. The method includes obtaining an observation characterizing a current state of an environment. For each layer parameter of each noisy layer of a neural network, a respective noise value is determined. For each layer parameter of each noisy layer, a noisy current value for the layer parameter is determined from a current value of the layer parameter, a current value of a corresponding noise parameter, and the noise value. A network input including the observation is processed using the neural network in accordance with the noisy current values to generate a network output for the network input. An action is selected from a set of possible actions to be performed by the agent in response to the observation using the network output.Type: ApplicationFiled: June 12, 2019Publication date: November 28, 2019Inventors: Olivier Pietquin, Jacob Lee Menick, Mohammad Gheshlaghi Azar, Bilal Piot, Volodymyr Mnih, Charles Blundell, Meire Fortunato, Remi Munos