In a method for improving converter and compiler operator granularity, a processor extracts an operator granularity from an artificial intelligence framework and an original model. A processor receives device characteristics from a target device. A processor outputs a converter granularity level to a converter based on the operator granularity and the device characteristics. A processor outputs a compiler granularity level to a compiler based on the operator granularity and the device characteristics.

A method of forming a neural network includes specifying layers of neural network neurons. A parameter genome is defined with numerical parameters characterizing connections between neural network neurons in the layers of neural network neurons, where the connections are defined from a neuron in a current layer to neurons in a set of adjacent layers, and where the parameter genome has a unique representation characterized by kilobytes of numerical parameters. Parameter genomes are combined into a connectome characterizing all connections between all neural network neurons in the connectome, where the connectome has in excess of millions of neural network neurons and billions of connections between the neural network neurons.

Systems and methods for tuning hyperparameters for a machine learning model using a challenger champion model are described. A set of challenger configurations are generated based on a hyperparameter for tuning and a subset of the set of challenger configurations are scheduled for evaluation based on a loss function. A loss value derived from the loss function for the challenger configurations is compared to a loss value derived from the loss function for a champion configuration, and the champion configuration is replaced with the challenger configuration based on the comparison of the loss value derived from the loss function for the challenger configuration and the loss value derived from the loss function for the champion configuration. When the champion is replaced, a new set of challenger configurations is generated based on the new champion configuration.

A processing system comprising multiple tiles and an interconnect between the tiles. The interconnect is used to communicate between a group of some or all of the tiles according to a bulk synchronous parallel scheme, whereby each tile in the group performs an on-tile compute phase followed by an inter-tile exchange phase with the exchange phase being held back until all tiles in the group have completed the compute phase. Each tile in the group has a local exit state upon completion of the compute phase. The instruction set comprises a synchronization instruction for execution by each tile upon completion of its compute phase to signal a sync request to logic in the interconnect. In response to receiving the sync request from all the tiles in the group, the logic releases the next exchange phase and also makes available an aggregated a state of all the tiles in the group.

The present invention is a water fountain control system that utilizes cameras to analyze movements of a human subject, and actuates one or more water fountain controllers in response to the movements to create a display incorporating spray patterns of the flowing water. The camera system records video in real time and generates optical signals that are sent to a processor running software that assesses the dimension, position, stance, and/or motion of the human subject and converts the data into recognized classes of movements and/or poses. Once the processor identifies the type of movements and/or poses, it sends signals to the actuators of the water fountains to control the fountains in a manner that implements stored predetermined visual effects generated by the fountain to create a visual presentation to an audience.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for generating an artificial neural network architecture based on a synaptic connectivity graph. According to one aspect, there is provided a method comprising: obtaining a synaptic resolution image of at least a portion of a brain of a biological organism; processing the image to identify: (i) a plurality of neurons in the brain, and (ii) a plurality of synaptic connections between pairs of neurons in the brain; generating data defining a graph representing synaptic connectivity between the neurons in the brain; determining an artificial neural network architecture corresponding to the graph representing the synaptic connectivity between the neurons in the brain; and processing a network input using an artificial neural network having the artificial neural network architecture to generate a network output.

Implementations are disclosed for automated design and implementation of machine learning (ML) state machines that include at least some aspect of machine learning. In various implementations, unstructured input may be received from a user. The unstructured input may convey operational aspect(s) of a machine learning (ML) state machine desired by the user. The unstructured input may be semantically processed to determine an intent of the user. The intent may include the operational aspect(s) of the ML state machine desired by the user. Based on the intent of the user, a plurality of modular logical routines may be selected from an existing library of modular logical routines. At least one logical routine of the selected plurality of logical routines may include logical operations that process data using one or more machine learning models. The selected plurality of logical routines may be assembled into the desired state ML state machine.

A processing system comprising an arrangement of tiles and an interconnect between the tiles. The interconnect comprises synchronization logic for coordinating a barrier synchronization to be performed between a group of the tiles. The instruction set comprises a synchronization instruction taking an operand which selects one of a plurality of available modes each specifying a different membership of the group. Execution of the synchronization instruction cause a synchronization request to be transmitted from the respective tile to the synchronization logic, and instruction issue to be suspended on the respective tile pending a synchronization acknowledgement being received back from the synchronization logic. In response to receiving the synchronization request from all the tiles in the group as specified by the operand of the synchronization instruction, the synchronization logic returns the synchronization acknowledgment to the tiles in the specified group.

A pre-training apparatus and method for reinforcement learning based on a Generative Adversarial Network (GAN) is provided. GAN includes a generator and a discriminator. The method comprising receiving training data from a real environment where the training data includes a data slice corresponding to a first state-reward pair and a first state-action pair, training the GAN using the training data, training a relations network to extract a latent relationship of the first state-action pair with the first state-reward pair in a reinforcement learning context, causing the generator trained with training data to generate first synthetic data, processing a portion of the first synthetic data in the relations network to generate a resulting data slice, merging the second state-action pair portion of the first synthetic data with the second state-reward pair from the relations network to generate second synthetic data to update a policy for interaction with the real environment.

In various examples, an adaptive eye tracking machine learning model engine (“adaptive-model engine”) for an eye tracking system is described. The adaptive-model engine may include an eye tracking or gaze tracking development pipeline (“adaptive-model training pipeline”) that supports collecting data, training, optimizing, and deploying an adaptive eye tracking model that is a customized eye tracking model based on a set of features of an identified deployment environment. The adaptive-model engine supports ensembling the adaptive eye tracking model that may be trained on gaze vector estimation in surround environments and ensemble based on a plurality of eye tracking variant models and a plurality of facial landmark neural network metrics.