Publications

Abstract not provided.

Abstract not provided.

With the amount of neuromorphic tools and frame-works growing in number, we recognize a need to increase interoperability within our field. As an illustration of this, we explore linking two independently constructed tools. Specifically, we detail the construction of an a execution backend based on STACS: Simulation Tool for Asynchronous Cortical Streams for the Fugu spiking neural algorithms framework. STACS extends the computational scope of Fugu, enabling fast simulation of large-scale neural networks. Combining these two tools is shown to be mutually beneficial, ultimately enabling more functionality than either tool on its own. We discuss design considerations, in-cluding recognizing the advantages of straightforward standards. Further, we provide some benchmark results showing drastic improvements in execution time.

Spiking Neural Networks (SNNs) have recently gained significant interest in on-chip learning in embedded devices and emerged as an energy-efficient alternative to conventional Artificial Neural Networks (ANNs). However, to extend SNNs to a Federated Learning (FL) setting involving collaborative model training, the communication between the local devices and the remote server remains the bottleneck, which is often restricted and costly. In this paper, we first explore the inherent robustness of SNNs under noisy communication in FL. Building upon this foundation, we propose a novel Federated Learning with Top-κ Sparsification (FLTS) algorithm to reduce the bandwidth usage for FL training. We discover that the proposed scheme with SNNs allows more bandwidth savings compared to ANNs without impacting the model's accuracy. Additionally, the number of parameters to be communicated can be reduced to as low as 6% of the size of the original model. We further improve the communication efficiency by enabling dynamic parameter compression during model training. Extensive experiment results demonstrate that our proposed algorithms significantly outperform the baselines in terms of communication cost and model accuracy and are promising for practical network-efficient FL with SNNs.

The lack of large, relevant and labeled datasets for synthetic aperture radar (SAR) automatic target recognition (ATR) poses a challenge for deep neural network approaches. In the case of SAR ATR, transfer learning offers promise where models are pre-trained on either synthetic SAR, alternatively collected SAR, or non-SAR source data and then fine-tuned on a smaller target SAR dataset. The concept being that the neural network can learn fundamental features from the more abundant source domain resulting in high accuracy and robust models when fine-tuned on a smaller target domain. One open question with this transfer learning strategy is how to choose source datasets that will improve accuracy of a target SAR dataset when the model is fine-tuned. Here, we apply a set of model and dataset transferability analysis techniques to investigate the efficacy of transfer learning for SAR ATR. In particular, we examine Optimal Transport Dataset Distance (OTDD), Log Maximum Evidence (LogMe), Log Expected Empirical Prediction (LEEP), Gaussian Bhattacharyya Coefficient (GBC), and H-Score. These methods consider properties such as task relatedness, statistical analysis of learned embedding properties, as well as distribution distances of the source and target domains. We apply these transferability metrics to ResNet18 models trained on a set of Non-SAR as well as SAR datasets. Overall, we present an investigation into quantitatively analyzing transferability for SAR ATR.

Deep neural networks for automatic target recognition (ATR) have been shown to be highly successful for a large variety of Synthetic Aperture Radar (SAR) benchmark datasets. However, the black box nature of neural network approaches raises concerns about how models come to their decisions, especially when in high-stake scenarios. Accordingly, a variety of techniques are being pursued seeking to offer understanding of machine learning algorithms. In this paper, we first provide an overview of explainability and interpretability techniques introducing their concepts and the insights they produce. Next we summarize several methods for computing specific approaches to explainability and interpretability as well as analyzing their outputs. Finally, we demonstrate the application of several attribution map methods and apply both attribution analysis metrics as well as localization interpretability analysis to six neural network models trained on the Synthetic and Measured Paired Labeled Experiment (SAMPLE) dataset to illustrate the insights these methods offer for analyzing SAR ATR performance.

With the amount of neuromorphic tools and frame-works growing in number, we recognize a need to increase interoperability within our field. As an illustration of this, we explore linking two independently constructed tools. Specifically, we detail the construction of an a execution backend based on STACS: Simulation Tool for Asynchronous Cortical Streams for the Fugu spiking neural algorithms framework. STACS extends the computational scope of Fugu, enabling fast simulation of large-scale neural networks. Combining these two tools is shown to be mutually beneficial, ultimately enabling more functionality than either tool on its own. We discuss design considerations, in-cluding recognizing the advantages of straightforward standards. Further, we provide some benchmark results showing drastic improvements in execution time.

Deep neural networks for automatic target recognition (ATR) have been shown to be highly successful for a large variety of Synthetic Aperture Radar (SAR) benchmark datasets. However, the black box nature of neural network approaches raises concerns about how models come to their decisions, especially when in high-stake scenarios. Accordingly, a variety of techniques are being pursued seeking to offer understanding of machine learning algorithms. In this paper, we first provide an overview of explainability and interpretability techniques introducing their concepts and the insights they produce. Next we summarize several methods for computing specific approaches to explainability and interpretability as well as analyzing their outputs. Finally, we demonstrate the application of several attribution map methods and apply both attribution analysis metrics as well as localization interpretability analysis to six neural network models trained on the Synthetic and Measured Paired Labeled Experiment (SAMPLE) dataset to illustrate the insights these methods offer for analyzing SAR ATR performance.

Co-design is a prominent topic presently in computing, speaking to the mutual benefit of coordinating design choices of several layers in the technology stack. For example, this may be designing algorithms that can most efficiently take advantage of the acceleration properties of a given architecture while simultaneously designing the hardware to support the structural needs of a class of computation. The implications of these design decisions are influential enough to be deemed a lottery, enabling an idea to win out over others irrespective of the individual merits. In this paper, we examine such research interactions through the lens of game theory. Our hope is that game theoretic analysis can provide greater insight into the decisions of neuromorphic co-design researchers and provide a formal argument that collaboration can be worth the cost. In particular, we consider the interplay between algorithm and architecture advances. The Colonel Blotto game is used to model and analyze different computing architectures, and the Stag Hunt model is used to analyze developments of spiking neural network algorithms and neuromorphic hardware as a co-design game. Our analysis illustrates challenges for either spiking algorithms or spiking architectures to advance the field independently. Instead, we find that cooperation can provide a much needed strategic advantage.

Evolutionary algorithms have been shown to be an effective method for training (or configuring) spiking neural networks. There are, however, challenges to developing accessible, scalable, and portable solutions. We present an extension to the Fugu framework that wraps the NEAT framework, bringing evolutionary algorithms to Fugu. This approach provides a flexible and customizable platform for optimizing network architectures, independent of fitness functions and input data structures. We leverage Fugu's computational graph approach to evaluate all members of a population in parallel. Additionally, as Fugu is platform-agnostic, this population can be evaluated in simulation or on neuromorphic hardware. We demonstrate our extension using several classification and agent-based tasks. One task illustrates how Fugu integration allows for spiking pre-processing to lower the search space dimensionality. We also provide some benchmark results using the Intel Loihi platform.

Abstract not provided.

Abstract not provided.

Finding the maximum cut of a graph (MAXCUT) is a classic optimization problem that has motivated parallel algorithm development. While approximate algorithms to MAXCUT offer attractive theoretical guarantees and demonstrate compelling empirical performance, such approximation approaches can shift the dominant computational cost to the stochastic sampling operations. Neuromorphic computing, which uses the organizing principles of the nervous system to inspire new parallel computing architectures, offers a possible solution. One ubiquitous feature of natural brains is stochasticity: the individual elements of biological neural networks possess an intrinsic randomness that serves as a resource enabling their unique computational capacities. By designing circuits and algorithms that make use of randomness similarly to natural brains, we hypothesize that the intrinsic randomness in microelectronics devices could be turned into a valuable component of a neuromorphic architecture enabling more efficient computations. Here, we present neuromorphic circuits that transform the stochastic behavior of a pool of random devices into useful correlations that drive stochastic solutions to MAXCUT. We show that these circuits perform favorably in comparison to software solvers and argue that this neuromorphic hardware implementation provides a path for scaling advantages. This work demonstrates the utility of combining neuromorphic principles with intrinsic randomness as a computational resource for new computational architectures.

Finding the maximum cut of a graph (MAXCUT) is a classic optimization problem that has motivated parallel algorithm development. While approximate algorithms to MAXCUT offer attractive theoretical guarantees and demonstrate compelling empirical performance, such approximation approaches can shift the dominant computational cost to the stochastic sampling operations. Neuromorphic computing, which uses the organizing principles of the nervous system to inspire new parallel computing architectures, offers a possible solution. One ubiquitous feature of natural brains is stochasticity: the individual elements of biological neural networks possess an intrinsic randomness that serves as a resource enabling their unique computational capacities. By designing circuits and algorithms that make use of randomness similarly to natural brains, we hypothesize that the intrinsic randomness in microelectronics devices could be turned into a valuable component of a neuromorphic architecture enabling more efficient computations. Here, we present neuromorphic circuits that transform the stochastic behavior of a pool of random devices into useful correlations that drive stochastic solutions to MAXCUT. We show that these circuits perform favorably in comparison to software solvers and argue that this neuromorphic hardware implementation provides a path for scaling advantages. This work demonstrates the utility of combining neuromorphic principles with intrinsic randomness as a computational resource for new computational architectures.

Though neuromorphic computers have typically targeted applications in machine learning and neuroscience (‘cognitive’ applications), they have many computational characteristics that are attractive for a wide variety of computational problems. In this work, we review the current state-of-the-art for non-cognitive applications on neuromorphic computers, including simple computational kernels for composition, graph algorithms, constrained optimization, and signal processing. We discuss the advantages of using neuromorphic computers for these different applications, as well as the challenges that still remain. The ultimate goal of this work is to bring awareness to this class of problems for neuromorphic systems to the broader community, particularly to encourage further work in this area and to make sure that these applications are considered in the design of future neuromorphic systems.

Abstract not provided.

Abstract not provided.

Neuromorphic computing (NMC) is an exciting paradigm seeking to incorporate principles from biological brains to enable advanced computing capabilities. Not only does this encompass algorithms, such as neural networks, but also the consideration of how to structure the enabling computational architectures for executing such workloads. Assessing the merits of NMC is more nuanced than simply comparing singular, historical performance metrics from traditional approaches versus that of NMC. The novel computational architectures require new algorithms to make use of their differing computational approaches. And neural algorithms themselves are emerging across increasing application domains. Accordingly, we propose following the example high performance computing has employed using context capturing mini-apps and abstraction tools to explore the merits of computational architectures. Here we present Neural Mini-Apps in a neural circuit tool called Fugu as a means of NMC insight.

Abstract not provided.

Abstract not provided.

Abstract not provided.

Neuromorphic computing, which aims to replicate the computational structure and architecture of the brain in synthetic hardware, has typically focused on artificial intelligence applications. What is less explored is whether such brain-inspired hardware can provide value beyond cognitive tasks. Here we show that the high degree of parallelism and configurability of spiking neuromorphic architectures makes them well suited to implement random walks via discrete-time Markov chains. These random walks are useful in Monte Carlo methods, which represent a fundamental computational tool for solving a wide range of numerical computing tasks. Using IBM’s TrueNorth and Intel’s Loihi neuromorphic computing platforms, we show that our neuromorphic computing algorithm for generating random walk approximations of diffusion offers advantages in energy-efficient computation compared with conventional approaches. We also show that our neuromorphic computing algorithm can be extended to more sophisticated jump-diffusion processes that are useful in a range of applications, including financial economics, particle physics and machine learning.

Abstract not provided.

Graph algorithms enable myriad large-scale applications including cybersecurity, social network analysis, resource allocation, and routing. The scalability of current graph algorithm implementations on conventional computing architectures are hampered by the demise of Moore’s law. We present a theoretical framework for designing and assessing the performance of graph algorithms executing in networks of spiking artificial neurons. Although spiking neural networks (SNNs) are capable of general-purpose computation, few algorithmic results with rigorous asymptotic performance analysis are known. SNNs are exceptionally well-motivated practically, as neuromorphic computing systems with 100 million spiking neurons are available, and systems with a billion neurons are anticipated in the next few years. Beyond massive parallelism and scalability, neuromorphic computing systems offer energy consumption orders of magnitude lower than conventional high-performance computing systems. We employ our framework to design and analyze new spiking algorithms for shortest path and dynamic programming problems. Our neuromorphic algorithms are message-passing algorithms relying critically on data movement for computation. For fair and rigorous comparison with conventional algorithms and architectures, which is challenging but paramount, we develop new models of data-movement in conventional computing architectures. This allows us to prove polynomial-factor advantages, even when we assume a SNN consisting of a simple grid-like network of neurons. To the best of our knowledge, this is one of the first examples of a rigorous asymptotic computational advantage for neuromorphic computing.

Abstract not provided.