Abstract: Disclosed is a data processing system that includes compile time logic to section a graph into a sequence of sections, configure a first section to generate a first set of output tiles in a first target tiling configuration in response to processing a first set of input tiles in a first input tiling configuration, and configure a second section to generate a second set of output tiles in a second target tiling configuration in response to processing the first set of output tiles in a second input tiling configuration. Runtime logic is configured to pad a first input into a first padded input, read the first set of input tiles from the first padded input in the first input tiling configuration, and process the first set of input tiles through the first section to generate the first set of output tiles in the first target tiling configuration.

Abstract: Disclosed is a data processing system that includes compile time logic configured to section a graph into a sequence of sections, and configure each section of the sequence of sections such that an input layer of a section processes an input, one or more intermediate layers of the corresponding section processes corresponding one or more intermediate outputs, and a final layer of the corresponding section generates a final output. The final output has a non-overlapping final tiling configuration, the one or more intermediate outputs have corresponding one or more overlapping intermediate tiling configurations, and the input has an overlapping input tiling configuration. The compile time logic is further to determine the various tiling configurations by starting from the final layer and reverse traversing through the one or more intermediate layers, and ending with the input layer.

Abstract: Techniques for training neural networks are provided. According to one set of embodiments, a first array is processed in a spreading component to produce a second array, where a first dimension of the first array corresponds to at least one sequence of approximately orthogonal numeric vectors representing tokens, and where the spreading component combines values along the first dimension. The second array is processed in a transformer neural network to determine correlations between the sequence, which produces a third array. One or more batches of the third array are processed in a de-spreading component to produce a fourth array.

Abstract: Embodiments of the present disclosure include a method for token-position handling comprising: processing a first sequence of tokens to produce a second sequence of tokens, wherein the second sequence of tokens has a smaller number of tokens than the first sequence of tokens; masking at least some tokens in the second sequence to produce masked tokens; moving the masked tokens to the beginning of the second sequence to produce a third sequence; encoding tokens in the third sequence into a set of numeric vectors in a first array; and processing the first array in a transformer neural network to determine correlations among the third sequence, the processing the first array producing a second array.

Abstract: A method including receiving, in a controller, from a client device in a network, a resolution query specifying a host name, is provided. The method includes parsing the resolution query to determine whether the host name is associated with an core host or with a public host, and directing the resolution query to a remote domain name system server dedicated to service the core host when the host name is associated with an enterprise name. The method also includes directing the resolution query to a local domain name system server when the host name is associated with a public service provided by the public host. A system to perform the above method is also provided.

Abstract: Methods for gradient accumulation with free momentum are performed by systems and devices during neural network model training. An accumulator that includes a processor circuit and a memory element generates free momentum between passes of a neural network model training process. The processor circuit receives a difference weight (gradient) and generates a first input by applying a weighting parameter thereto. The processor circuit obtains a prior weight from the memory element and generates a second input by applying another weighting parameter thereto. The processor circuit generates a filtered input with momentum by filtering the first and second input. The memory element generates a stored next pass weight by accumulating the filtered input with the prior weight. A computing resource then processes the next pass of the neural network model training using the stored next pass weight. The methods, systems, and devices are applicable to pipelined model parallelism training processes.

Abstract: Improved convolutional neural network-based machine learning models are disclosed herein. A convolutional neural network is configured to decompose feature maps generated based on a data item to be classified. The feature maps are decomposed into a first and second subsets. The first subset is representative of high frequency components of the data item, and the second subset is representative of low frequency components of the data item. The second subset is upsampled and is combined with the first subset. The combined feature maps are convolved with a filter to extract a set of features associated with the data item. The first subset is also downsampled and combined with the second subset. The combined feature maps are convolved with a filter to extract another set of features. The data item is classified based on the sets of features extracted based on the convolution operations.

Abstract: A method including receiving, in a controller, from a client device in a network, a resolution query specifying a host name, is provided. The method includes parsing the resolution query to determine whether the host name is associated with an core host or with a public host, and directing the resolution query to a remote domain name system server dedicated to service the core host when the host name is associated with an enterprise name. The method also includes directing the resolution query to a local domain name system server when the host name is associated with a public service provided by the public host. A system to perform the above method is also provided.

Abstract: A method and system to generate a probabilistic prediction of the presence/absence of cancer in longitudinal and current image datasets, and/or multimodal image datasets, and the location of the cancer, is described. The method and system uses an ensemble of deep learning models. The ensemble includes a global model in the form of a 3D convolutional neural network (CNN) extracting features in the datasets indicative of the presence of cancer on a global basis. The ensemble also includes a two-stage prediction model which includes a first stage or detection model which identifies cancer detection candidates (different cropped volumes of 3D data in the a dataset containing candidates which may be cancer) and a second stage or probability model which incorporates the longitudinal datasets (or multimodal images in a multimodal dataset) and the extracted features from the global model and assigns a cancer probability p to each of the cancer detection candidates.

Abstract: Systems and methods for pipelined neural network processing with continuous and asynchronous updates are described. A method for processing a neural network comprising L layers, where L is an integer greater than two, includes partitioning the L layers among a set of computing resources configured to process forward passes and backward passes associated with each of the L layers. The method further includes initiating processing of the forward passes and the backward passes using the set of computing resources. The method further includes upon completion of a first set of forward passes and a first set of backward passes associated with a first layer of the L layers, initiating update of parameters associated with the first layer when gradients are available for updating the parameters associated with the first layer without waiting to calculate gradients associated with any of remaining L layers.

Abstract: Methods, systems, apparatuses, and computer program products are described herein that enable execution of a large AI model on a memory-constrained target device that is communicatively connected to a parameter server, which stores a master copy of the AI model. The AI model may be dissected into smaller portions (e.g., layers or sub-layers), and each portion may be executed as efficiently as possible on the target device. After execution of one portion of the AI model is finished, another portion of the AI model may be downloaded and executed at the target device. To improve efficiency, the input samples may be divided into microbatches, and a plurality of microbatches executing in sequential order may form a minibatch. The size of the group of microbatches or minibatch can be adjusted to reduce the communication overhead. Multi-level parallel parameters reduction may be performed at the parameter server and the target device.

Abstract: Methods, systems, apparatuses, and computer program products are described herein that enable execution of a large AI model on a memory-constrained target device that is communicatively connected to a parameter server, which stores a master copy of the AI model. The AI model may be dissected into smaller portions (e.g., layers or sub-layers), and each portion may be executed as efficiently as possible on the target device. After execution of one portion of the AI model is finished, another portion of the AI model may be downloaded and executed at the target device. This paradigm of executing one portion of the AI model at a time allows for dynamic execution of the large AI model.

Abstract: A method including receiving, in a controller, from a client device in a network, a resolution query specifying a host name, is provided. The method includes parsing the resolution query to determine whether the host name is associated with an core host or with a public host, and directing the resolution query to a remote domain name system server dedicated to service the core host when the host name is associated with an enterprise name. The method also includes directing the resolution query to a local domain name system server when the host name is associated with a public service provided by the public host. A system to perform the above method is also provided.

Abstract: A device having a sheath, a medical device, a stylet, a handle having a chamber portion and a plunger device. The handle connects to proximal ends of the sheath, the medical device, and the stylet. The chamber portion includes a volume of space configured to volumetrically connect to at a lumen of the medical device, the sheath or the stylet. The plunger device slides and moves within the chamber portion. The plunger device is able to pneumatically or hydraulically isolate a proximal portion of the volume of space from a distal portion of the volume of space. Proximal movement of the stylet and the plunger device cause a suction effect (i.e., negative or reduced pressure) at the distal end of the sheath and/or the medical device.

Abstract: A non-nematicidal composition comprises at least one glucan and/or at least one fucan which act individually or synergistically with mannitol to reduce losses in crop yield and marketable grade caused by the infestation of growth media with plant pathogenic nematodes, to levels equivalent to those achieved with commercial nematicides, but without posing a risk to the ecosystem or user. In some cases the composition comprises at least one glucan, at least one fucan and at least one mannitol which may be in a weight/weight ratio of approximately 1:2:3 of at least one glucan:at least one fucan:at least one mannitol or between approximately 1:1:1 to 1:1:3 of at least one glucan:at least one fucan:at least one mannitol.

Abstract: Disclosed embodiments include apparatuses, systems, and methods for providing an atraumatic sheath tip. Various disclosed embodiments seek to help reduce or avoid unnecessary tissue damage upon a sheath being extended to convey an elongated instrument for sampling or treatment. In an illustrative embodiment, an apparatus includes a deformable sheath tip configured to be positioned at a distal end of a sheath. The sheath defines therein a lumen configured to convey an elongated instrument. The sheath tip has a base end disposable at the distal end of the sheath and a contact end. The sheath tip has a first column strength along an axis of the sheath tip that is less than a second column strength of the sheath and a first degree of deformability that is greater than a second degree of deformability of the sheath.

Abstract: A fistula creation system for quickly creating a fistula between a first biological tissue and a second biological tissue of a body. The system includes a tissue fastening tool for applying pressure to a periphery around an area of a first biological tissue and a second biological tissue of a body, a power source and a fistula creating device. The fistula creating device includes one or more electrodes located at the distal end of a catheter and are in electrical communication with the power source. The electrodes cut the first and second biological tissue within the area when charged by the power source. The fistula creating device includes a balloon with an electrode for expanding the size of the fistula.

Abstract: A non-nematicidal composition comprises at least one glucan and/or at least one fucan which act individually or synergistically with mannitol to reduce losses in crop yield and marketable grade caused by the infestation of growth media with plant pathogenic nematodes, to levels equivalent to those achieved with commercial nematicides, but without posing a risk to the ecosystem or user. In some cases the composition comprises at least one glucan, at least one fucan and at least one mannitol which may be in a weight/weight ratio of approximately 1:2:3 of at least one glucan:at least one fucan:at least one mannitol or between approximately 1:1:1 to 1:1:3 of at least one glucan:at least one fucan:at least one mannitol.