Preferred Networks, Inc.

An inference apparatus includes one or more memories; and one or more processors. The one or more processors configured to acquire a latent state from input data regarding a control target; acquire a future latent state from the latent state and control data; infer, from the future latent state, a time series of a task to be executed by the control target to be controlled based on the control data; calculate a loss between the time series of the task and data indicating a target state; and update the control data based on the loss.

An information processing system includes at least one memory, and at least one processor. The at least one processor is configured to acquire one or more pieces of reference information used for generating output information based on user information, generate input information to a machine learning model based on the one or more pieces of reference information, and generate the output information by inputting the input information to the machine learning model.

A server device configured to communicate, via a communication network, with at least one device including a learner configured to perform processing by using a learned model, includes processor, a transmitter, and a storage configured to store a plurality of shared models pre-learned in accordance with environments and conditions of various devices. The processor is configured to acquire device data including information on an environment and conditions from the at least one device, and select an optimum shared model for the at least one device based on the acquired device data. The transmitter is configured to transmit a selected shared model to the at least one device.

A server device includes one or more memories and processors. The one or more memories hold one or more reconstruction models for generating a time series of free-viewpoint images, having been trained in advance to reconstruct a scene by using a time series of captured images from a plurality of viewpoints, obtained by capturing the scene from each of the plurality of viewpoints continuously in time. The one or more processors receive a request including viewpoint and time information for the scene from a dedicated application or a browser; generate, by using the one or more reconstruction models, the time series of free-viewpoint images corresponding to the viewpoint and time information included in the received request; and transmit, to the dedicated application or the browser having transmitted the request, the generated time series of free-viewpoint images in a video format that is supported by the dedicated application or the browser.

An information processing device includes at least one memory and at least one processor. The at least one processor is configured to acquire information about a search target including at least information on a chemical element from another information processing device, repeatedly execute acquisition of a plurality of structures of a substance each containing the chemical element, update of the plurality of structures of the substance, and evaluation of the updated plurality of structures of the substance using a neural network, until a predetermined condition is satisfied, and after the repeated execution, transmit information about a searched structure of a substance and information about an evaluation value of the searched structure of the substance to the other information processing device. The at least one processor executes the acquisition based on the evaluation in the repeated execution.

A processor includes an arithmetic circuit configured to execute an arithmetic instruction; and a register configured to hold data used by the arithmetic circuit. The processor receives a data movement instruction and the arithmetic instruction corresponding to the data movement instruction, and moves the data from a first memory to the register based on a data movement instruction. The arithmetic circuit executes the arithmetic instruction after a data movement of the data movement instruction is completed.

An information processing system includes a plurality of memories and a plurality of processors configured to perform parallel processing using a model. The plurality of processors execute communication processing of a result of executing computational processing using at least a part of the model for first input data, and computational processing using at least a part of the model for second input data, such that processing periods of the communication processing and the computational processing at least partially overlap.

An arithmetic processing device includes second blocks, each including first blocks and one second memory, and each of the first blocks including one arithmetic unit and one first memory. The arithmetic processing device performs, in parallel, at least one of first, second, third, or fourth data transfers, by executing an instruction sequence. Sources and destinations of the first data transfers are one or more first blocks, sources of the second data transfers are one or more first blocks, destinations thereof are one or more second blocks, sources of the third data transfers are one or more second blocks, destinations thereof are one or more first blocks, and sources and destinations of the fourth data transfers are one or more second blocks. The instruction sequence includes a combination and execution order of at least one multicast instruction selected from more than one type of multicast instructions.

A processing system includes a first chip including one or more first processors; and a second chip including one or more second processors. For a weight of a neural network, the one or more second processors execute a backward operation of the neural network and calculate a gradient of the weight, the calculated gradient is transferred to the one or more first processors, and the one or more first processors update the weight based on the calculated gradient. The one or more first processors transfer the updated weight to the second chip. The one or more first processors update the weight of the neural network in accordance with a method using multiple types of parameters, the one or more first processors updating two or more types of parameters to update the weight, and the multiple types of parameters including the weight and the two or more types of parameters.

A device includes a memory and a processor. The processor is configured to acquire a first structure including information on an execution order of instructions and information on an amount of memory required for execution of each instruction, generate a second structure in which the information on the execution order and the information on the amount of memory are updated, by executing a manipulation on the first structure, acquire an evaluation value based on the second structure, and update the first structure using the second structure where the evaluation value satisfies a first condition. The first structure and the second structure are structures in which change and value calculation for a predetermined interval are supported, and the processor repeats the generation of the second structure, the acquisition of the evaluation value, and the update of the first structure when the evaluation value satisfies the first condition.