Patents by Inventor Sam Michael DEVLIN
Sam Michael DEVLIN 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: 20230405468Abstract: Aspects of the present disclosure provide systems and methods which utilizes machine learning techniques to provide enhanced accessibility features to a game. An accessibility service is provided which is capable of instantiating one or more machine learning models which can process current gameplay states and generate commands to assist users during gameplay. The accessibility commands may be provided to a game and used to supplement or modify user provided inputs in order to compensate for specific user needs. In further aspects, an accessibility user interface is provided which allows a user to dynamically enable or disable accessibility features during gameplay. The user interface is operable to receive accessibility selections and provide the selection data to an accessibility service during gameplay.Type: ApplicationFiled: May 19, 2023Publication date: December 21, 2023Applicant: Microsoft Technology Licensing, LLCInventors: Christopher John BROCKETT, Gabriel A. DESGARENNES, Sudha RAO, Hamid PALANGI, Ryan VOLUM, Yun Hui XU, Sam Michael DEVLIN, Brannon J. ZAHAND
-
Publication number: 20230111659Abstract: An apparatus has a memory storing a reinforcement learning policy with an optimization component and a data collection component. The apparatus has a regularization component which applies regularization selectively between the optimization component of the reinforcement learning policy and the data collection component of the reinforcement learning policy. A processor carries out a reinforcement learning process by: triggering execution of an agent according to the policy and with respect to a first task; observing values of variables comprising: an observation space of the agent, an action of the agent; and updating the policy using reinforcement learning according to the observed values and taking into account the regularization.Type: ApplicationFiled: November 2, 2022Publication date: April 13, 2023Inventors: Sam Michael DEVLIN, Maximilian IGL, Kamil Andrzej CIOSEK, Yingzhen LI, Sebastian TSCHIATSCHEK, Cheng ZHANG, Katja HOFMANN
-
Patent number: 11526812Abstract: An apparatus has a memory storing a reinforcement learning policy with an optimization component and a data collection component. The apparatus has a regularization component which applies regularization selectively between the optimization component of the reinforcement learning policy and the data collection component of the reinforcement learning policy. A processor carries out a reinforcement learning process by: triggering execution of an agent according to the policy and with respect to a first task; observing values of variables comprising: an observation space of the agent, an action of the agent; and updating the policy using reinforcement learning according to the observed values and taking into account the regularization.Type: GrantFiled: October 1, 2019Date of Patent: December 13, 2022Assignee: Microsoft Technology Licensing, LLCInventors: Sam Michael Devlin, Maximilian Igl, Kamil Andrzej Ciosek, Yingzhen Li, Sebastian Tschiatschek, Cheng Zhang, Katja Hofmann
-
Publication number: 20220114474Abstract: In various examples there is a method performed by a computer-implemented agent in an environment. The method comprises storing a reinforcement learning policy for controlling the computer-implemented agent. The method also comprises storing a distribution as a latent representation of a belief of the computer-implemented agent about at least one other agent in the environment. The method involves executing the computer-implemented agent according to the policy conditioned on parameters characterizing the distribution.Type: ApplicationFiled: October 9, 2020Publication date: April 14, 2022Inventors: Katja HOFMANN, Luisa Maria ZINTGRAF, Sam Michael DEVLIN, Kamil Andrzej CIOSEK
-
Patent number: 11182698Abstract: An apparatus is described for training a behavior of an agent in a physical or digital environment. The apparatus comprises a memory storing the location of at least one reward token in the environment. The location has been specified by a user. At least one processor executes the agent in the environment according to a behavior policy. The processor is configured to observe values of variables comprising: an observation of the agent, an action of the agent and any reward resulting from the reward token. The processor is configured to update the behavior policy using reinforcement learning according to the observed values.Type: GrantFiled: July 10, 2019Date of Patent: November 23, 2021Assignee: Microsoft Technology Licensing, LLCInventors: Anthony David Joseph Diggle, Jay Nanavati, Katja Hofmann, Sam Michael Devlin, Andrew Philip Slowey, Janhavi Agrawal, David Michael Bignell, Adrian Kieron O'Grady
-
Publication number: 20210097445Abstract: An apparatus has a memory storing a reinforcement learning policy with an optimization component and a data collection component. The apparatus has a regularization component which applies regularization selectively between the optimization component of the reinforcement learning policy and the data collection component of the reinforcement learning policy. A processor carries out a reinforcement learning process by: triggering execution of an agent according to the policy and with respect to a first task; observing values of variables comprising: an observation space of the agent, an action of the agent; and updating the policy using reinforcement learning according to the observed values and taking into account the regularization.Type: ApplicationFiled: October 1, 2019Publication date: April 1, 2021Inventors: Sam Michael DEVLIN, Maximilian IGL, Kamil Andrzej CIOSEK, Yingzhen LI, Sebastian TSCHIATSCHEK, Cheng ZHANG, Katja HOFMANN
-
Publication number: 20200356897Abstract: An apparatus is described for training a behavior of an agent in a physical or digital environment. The apparatus comprises a memory storing the location of at least one reward token in the environment. The location has been specified by a user. At least one processor executes the agent in the environment according to a behavior policy. The processor is configured to observe values of variables comprising: an observation of the agent, an action of the agent and any reward resulting from the reward token. The processor is configured to update the behavior policy using reinforcement learning according to the observed values.Type: ApplicationFiled: July 10, 2019Publication date: November 12, 2020Inventors: Anthony David Joseph DIGGLE, Jay NANAVATI, Katja HOFMANN, Sam Michael DEVLIN, Andrew Philip SLOWEY, Janhavi AGRAWAL, David Michael BIGNELL, Adrian Kieron O'GRADY
-
Patent number: 10708323Abstract: Systems for managing content in a cloud-based service platform. A server in a cloud-based environment is interfaced with storage devices that hold one or more stored objects accessible by two or more users. The stored objects comprise folders and files as well as other objects such as workflow objects that are associated with the folders or the files. The workflow objects comprise workflow metadata that describes a workflow as a set of workflow tasks to be carried out in a progression. Processing of a workflow task and/or carrying out a portion of the progression includes modification of shared content objects. The processing or modification events are detected through workflow events, which in turn cause one or more workflow responses to be generated. Workflow responses comprise updates to the workflow metadata to record progression through the workflow and/or workflow responses comprise updates to any one or more of the stored objects.Type: GrantFiled: July 30, 2018Date of Patent: July 7, 2020Assignee: Box, Inc.Inventors: Anne Elizabeth Hiatt Pearl, Jenica Nash Blechschmidt, Natalia Vinnik, Robert Kyle Waldrop, Sam Michael Devlin, Steven Luis Cipolla, Sesh Jalagam
-
Publication number: 20180337966Abstract: Systems for managing content in a cloud-based service platform. A server in a cloud-based environment is interfaced with storage devices that hold one or more stored objects accessible by two or more users. The stored objects comprise folders and files as well as other objects such as workflow objects that are associated with the folders or the files. The workflow objects comprise workflow metadata that describes a workflow as a set of workflow tasks to be carried out in a progression. Processing of a workflow task and/or carrying out a portion of the progression includes modification of shared content objects. The processing or modification events are detected through workflow events, which in turn cause one or more workflow responses to be generated. Workflow responses comprise updates to the workflow metadata to record progression through the workflow and/or workflow responses comprise updates to any one or more of the stored objects.Type: ApplicationFiled: July 30, 2018Publication date: November 22, 2018Applicant: Box, Inc.Inventors: Anne Elizabeth Hiatt PEARL, Jenica Nash BLECHSCHMIDT, Natalia VINNIK, Robert Kyle WALDROP, Sam Michael DEVLIN, Steven Luis CIPOLLA, Sesh JALAGAM
-
Patent number: 10038731Abstract: Systems for managing content in a cloud-based service platform. A server in a cloud-based environment is interfaced with storage devices that hold one or more stored objects accessible by two or more users. The stored objects comprise folders and files as well as other objects such as workflow objects that are associated with the folders or the files. The workflow objects comprise workflow metadata that describes a workflow as a set of workflow tasks to be carried out in a progression. Processing of a workflow task and/or carrying out a portion of the progression includes modification of shared content objects. The processing or modification events are detected through workflow events, which in turn cause one or more workflow responses to be generated. Workflow responses comprise updates to the workflow metadata to record progression through the workflow and/or workflow responses comprise updates to any one or more of the stored objects.Type: GrantFiled: August 22, 2016Date of Patent: July 31, 2018Assignee: Box, Inc.Inventors: Anne Elizabeth Hiatt Pearl, Jenica Nash Blechschmidt, Natalia Vinnik, Robert Kyle Waldrop, Sam Michael Devlin, Steven Luis Cipolla, Sesh Jalagam
-
Publication number: 20170048285Abstract: Systems for managing content in a cloud-based service platform. A server in a cloud-based environment is interfaced with storage devices that hold one or more stored objects accessible by two or more users. The stored objects comprise folders and files as well as other objects such as workflow objects that are associated with the folders or the files. The workflow objects comprise workflow metadata that describes a workflow as a set of workflow tasks to be carried out in a progression. Processing of a workflow task and/or carrying out a portion of the progression includes modification of shared content objects. The processing or modification events are detected through workflow events, which in turn cause one or more workflow responses to be generated. Workflow responses comprise updates to the workflow metadata to record progression through the workflow and/or workflow responses comprise updates to any one or more of the stored objects.Type: ApplicationFiled: August 22, 2016Publication date: February 16, 2017Applicant: Box, Inc.Inventors: Anne Elizabeth Hiatt PEARL, Jenica Nash BLECHSCHMIDT, Natalia VINNIK, Robert Kyle WALDROP, Sam Michael DEVLIN, Steven Luis CIPOLLA, Sesh JALAGAM