公开(公告)号：US20210110115A1

公开(公告)日：2021-04-15

申请号：US16497602

申请日：2018-06-05

Applicant: DeepMind Technologies Limited

Inventor： Karl Moritz Hermann ,  Philip Blunsom ,  Felix George Hill

IPC: G06F40/30 ,  G06N3/08 ,  G06F17/16

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting actions to be performed by an agent interacting with an environment. In one aspect, a system includes a language encoder model that is configured to receive a text string in a particular natural language, and process the text string to generate a text embedding of the text string. The system includes an observation encoder neural network that is configured to receive an observation characterizing a state of the environment, and process the observation to generate an observation embedding of the observation. The system includes a subsystem that is configured to obtain a current text embedding of a current text string and a current observation embedding of a current observation. The subsystem is configured to select an action to be performed by the agent in response to the current observation.

公开(公告)号：US11354509B2

公开(公告)日：2022-06-07

申请号：US16497602

申请日：2018-06-05

Applicant: DeepMind Technologies Limited

Inventor： Karl Moritz Hermann ,  Philip Blunsom ,  Felix George Hill

IPC: G06F40/30 ,  G06F17/16 ,  G06N3/08

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting actions to be performed by an agent interacting with an environment. In one aspect, a system includes a language encoder model that is configured to receive a text string in a particular natural language, and process the text string to generate a text embedding of the text string. The system includes an observation encoder neural network that is configured to receive an observation characterizing a state of the environment, and process the observation to generate an observation embedding of the observation. The system includes a subsystem that is configured to obtain a current text embedding of a current text string and a current observation embedding of a current observation. The subsystem is configured to select an action to be performed by the agent in response to the current observation.

公开(公告)号：US20200005147A1

公开(公告)日：2020-01-02

申请号：US16565245

申请日：2019-09-09

Applicant: DeepMind Technologies Limited

Inventor： Edward Thomas Grefenstette ,  Karl Moritz Hermann ,  Mustafa Suleyman ,  Philip Blunsom

IPC: G06N3/08 ,  G06N3/04 ,  G06N3/063

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for augmenting neural networks with an external memory. One of the methods includes providing an output derived from the neural network output for the time step as a system output for the time step; maintaining a current state of the external memory; determining, from the neural network output for the time step, memory state parameters for the time step; updating the current state of the external memory using the memory state parameters for the time step; reading data from the external memory in accordance with the updated state of the external memory; and combining the data read from the external memory with a system input for the next time step to generate the neural network input for the next time step.

公开(公告)号：US20220318516A1

公开(公告)日：2022-10-06

申请号：US17744921

申请日：2022-05-16

Applicant: DeepMind Technologies Limited

Inventor： Karl Moritz Hermann ,  Philip Blunsom ,  Felix George Hill

IPC: G06F40/30 ,  G06F17/16 ,  G06N3/08

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting actions to be performed by an agent interacting with an environment. In one aspect, a system includes a language encoder model that is configured to receive a text string in a particular natural language, and process the text string to generate a text embedding of the text string. The system includes an observation encoder neural network that is configured to receive an observation characterizing a state of the environment, and process the observation to generate an observation embedding of the observation. The system includes a subsystem that is configured to obtain a current text embedding of a current text string and a current observation embedding of a current observation. The subsystem is configured to select an action to be performed by the agent in response to the current observation.

公开(公告)号：US12265795B2

公开(公告)日：2025-04-01

申请号：US18649774

申请日：2024-04-29

Applicant: DeepMind Technologies Limited

Inventor： Karl Moritz Hermann ,  Philip Blunsom ,  Felix George Hill

IPC: G06F40/30 ,  G06F17/16 ,  G06N3/08

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting actions to be performed by an agent interacting with an environment. In one aspect, a system includes a language encoder model that is configured to receive a text string in a particular natural language, and process the text string to generate a text embedding of the text string. The system includes an observation encoder neural network that is configured to receive an observation characterizing a state of the environment, and process the observation to generate an observation embedding of the observation. The system includes a subsystem that is configured to obtain a current text embedding of a current text string and a current observation embedding of a current observation. The subsystem is configured to select an action to be performed by the agent in response to the current observation.

公开(公告)号：US20240320438A1

公开(公告)日：2024-09-26

申请号：US18649774

申请日：2024-04-29

Applicant: DeepMind Technologies Limited

Inventor： Karl Moritz Hermann ,  Philip Blunsom ,  Felix George Hill

IPC: G06F40/30 ,  G06F17/16 ,  G06N3/08

CPC classification number: G06F40/30 ,  G06F17/16 ,  G06N3/08

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting actions to be performed by an agent interacting with an environment. In one aspect, a system includes a language encoder model that is configured to receive a text string in a particular natural language, and process the text string to generate a text embedding of the text string. The system includes an observation encoder neural network that is configured to receive an observation characterizing a state of the environment, and process the observation to generate an observation embedding of the observation. The system includes a subsystem that is configured to obtain a current text embedding of a current text string and a current observation embedding of a current observation. The subsystem is configured to select an action to be performed by the agent in response to the current observation.

公开(公告)号：US12008324B2

公开(公告)日：2024-06-11

申请号：US17744921

申请日：2022-05-16

Applicant: DeepMind Technologies Limited

Inventor： Karl Moritz Hermann ,  Philip Blunsom ,  Felix George Hill

IPC: G06F40/30 ,  G06F17/16 ,  G06N3/08

CPC classification number: G06F40/30 ,  G06F17/16 ,  G06N3/08

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for selecting actions to be performed by an agent interacting with an environment. In one aspect, a system includes a language encoder model that is configured to receive a text string in a particular natural language, and process the text string to generate a text embedding of the text string. The system includes an observation encoder neural network that is configured to receive an observation characterizing a state of the environment, and process the observation to generate an observation embedding of the observation. The system includes a subsystem that is configured to obtain a current text embedding of a current text string and a current observation embedding of a current observation. The subsystem is configured to select an action to be performed by the agent in response to the current observation.

公开(公告)号：US11593640B2

公开(公告)日：2023-02-28

申请号：US16565245

申请日：2019-09-09

Applicant: DeepMind Technologies Limited

Inventor： Edward Thomas Grefenstette ,  Karl Moritz Hermann ,  Mustafa Suleyman ,  Philip Blunsom

IPC: G06N3/08 ,  G06N3/04 ,  G06N3/063 ,  G06N3/082

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for augmenting neural networks with an external memory. One of the methods includes providing an output derived from the neural network output for the time step as a system output for the time step; maintaining a current state of the external memory; determining, from the neural network output for the time step, memory state parameters for the time step; updating the current state of the external memory using the memory state parameters for the time step; reading data from the external memory in accordance with the updated state of the external memory; and combining the data read from the external memory with a system input for the next time step to generate the neural network input for the next time step.

公开(公告)号：US12099928B2

公开(公告)日：2024-09-24

申请号：US18174394

申请日：2023-02-24

Applicant: DeepMind Technologies Limited

Inventor： Edward Thomas Grefenstette ,  Karl Moritz Hermann ,  Mustafa Suleyman ,  Philip Blunsom

IPC: G06N3/08 ,  G06N3/044 ,  G06N3/063 ,  G06N3/082

CPC classification number: G06N3/08 ,  G06N3/044 ,  G06N3/063 ,  G06N3/082

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for augmenting neural networks with an external memory. One of the methods includes providing an output derived from the neural network output for the time step as a system output for the time step; maintaining a current state of the external memory; determining, from the neural network output for the time step, memory state parameters for the time step; updating the current state of the external memory using the memory state parameters for the time step; reading data from the external memory in accordance with the updated state of the external memory; and combining the data read from the external memory with a system input for the next time step to generate the neural network input for the next time step.

公开(公告)号：US20230289598A1

公开(公告)日：2023-09-14

申请号：US18174394

申请日：2023-02-24

Applicant: DeepMind Technologies Limited

Inventor： EDWARD THOMAS GREFENSTETTE ,  Karl Moritz Hermann ,  Mustafa Suleyman ,  Philip Blunsom

IPC: G06N3/08 ,  G06N3/063 ,  G06N3/044 ,  G06N3/082

CPC classification number: G06N3/08 ,  G06N3/063 ,  G06N3/044 ,  G06N3/082

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for augmenting neural networks with an external memory. One of the methods includes providing an output derived from the neural network output for the time step as a system output for the time step; maintaining a current state of the external memory; determining, from the neural network output for the time step, memory state parameters for the time step; updating the current state of the external memory using the memory state parameters for the time step; reading data from the external memory in accordance with the updated state of the external memory; and combining the data read from the external memory with a system input for the next time step to generate the neural network input for the next time step.