Explore the vast range of titles available.

Spiking Neural Networks are often touted as brain-inspired learning models for the third wave of Artificial Intelligence. Although recent SNNs trained with supervised backpropagation show classification accuracy comparable to deep networks, the performance of unsupervised learning-based SNNs remains much lower. This paper presents a heterogeneous recurrent spiking neural network (HRSNN) with unsupervised learning for spatio-temporal classification of video activity recognition tasks on RGB (KTH, UCF11, UCF101) and event-based datasets (DVS128 Gesture). We observed an accuracy of 94.32% for the KTH dataset, 79.58% and 77.53% for the UCF11 and UCF101 datasets, respectively, and an accuracy of 96.54% on the event-based DVS Gesture dataset using the novel unsupervised HRSNN model. The key novelty of the HRSNN is that the recurrent layer in HRSNN consists of heterogeneous neurons with varying firing/relaxation dynamics, and they are trained via heterogeneous spike-time-dependent-plasticity (STDP) with varying learning dynamics for each synapse. We show that this novel combination of heterogeneity in architecture and learning method outperforms current homogeneous spiking neural networks. We further show that HRSNN can achieve similar performance to state-of-the-art backpropagation trained supervised SNN, but with less computation (fewer neurons and sparse connection) and less training data.

A Spiking Neural Network (SNN) is trained with Spike Timing Dependent Plasticity (STDP), which is a neuro-inspired unsupervised learning method for various machine learning applications. This paper studies the generalizability properties of the STDP learning processes using the Hausdorff dimension of the trajectories of the learning algorithm. The paper analyzes the effects of STDP learning models and associated hyper-parameters on the generalizability properties of an SNN. The analysis is used to develop a Bayesian optimization approach to optimize the hyper-parameters for an STDP model for improving the generalizability properties of an SNN.

We present a processing-in-memory (PIM)-based hardware platform, referred to as MONETA, for on-chip acceleration of inference and learning in hybrid convolutional spiking neural network. MONETA uses 8T SRAM-based PIM cores for vector matrix multiplication (VMM) augmented with Spike-Time-Dependent-Plasticity (STDP) based weight update. The SNN-focused data flow is presented to minimize data movement in MONETA while ensuring learning accuracy. MONETA supports on-line and on-chip training on PIM architecture. The STDP-trained ConvSNN with the proposed data flow, 4-bit input precision and 8-bit weight precision shows only 1.63 % lower accuracy in CIFAR-10 compared to the STDP accuracy implemented by the software. Further, the proposed architecture is used to accelerate a hybrid SNN architecture that couples off-chip supervised (back propagation through time) and on-chip unsupervised (STDP) training. We also evaluate the hybrid network architecture with the proposed data flow. The accuracy of this hybrid network is 11.58% higher than STDP trained accuracy result and 1.40 % higher comparing to the backpropaged training based ConvSNN result. Physical design of MONETA in 65nm CMOS shows 18.69 TOPS/W, 7.25 TOPS/W and 10.41 TOPS/W power efficiencies for the inference mode, learning mode and hybrid learning mode, respectively.

Artificial intelligence is at a crossroads: conventional deep learning models, while powerful, remain fundamentally limited in their ability to process information in time, adapt seamlessly to changing environments, and efficiently encode structured memory. The brain, by contrast, operates through spikes—discrete, event-driven signals that inherently capture temporal dependencies. Spiking Neural Networks (SNNs) have long been viewed primarily as energy-efficient alternatives to artificial neural networks (ANNs). This dissertation takes a fundamentally different view: it positions SNNs as a new computational paradigm, capable of expressing forms of temporal reasoning and adaptive intelligence that conventional deep learning struggles to achieve.However, realizing this vision requires overcoming a key limitation—most existing SNN models are built on homogeneous neuron and synapse dynamics, constraining their expressivity and adaptability. From a dynamical systems perspective, this homogeneity forces all neurons to evolve along similar timescales, reducing the network’s ability to capture multi-scale dependencies and limiting the richness of its attractor landscape. By contrast, in complex dynamical systems—including the brain—heterogeneous timescales create diverse trajectories in state space, enhancing stability, memory capacity, and computational flexibility. To address this, this dissertation introduces Heterogeneous Recurrent Spiking Neural Networks (HRSNNs), a novel class of SNNs that leverage diverse neuronal and synaptic timescales to improve learning efficiency and temporal representation. By incorporating heterogeneity, HRSNNs enable structured memory retention, greater robustness to non-stationary inputs, and improved real-time adaptability.Yet, heterogeneity alone is insufficient to fully harness the computational power of SNNs. To systematically extract their unique advantages, this dissertation develops a new mathematical framework that bridges spike-based processing with dynamical systems theory, state-space models (SSMs), and Lyapunov stability analysis. These tools provide formal guarantees on stability, convergence, and learning efficiency, key properties that have remained elusive in SNN research. Additionally, this work proposes a task-agnostic pruning methodology, which sparsifies SNNs not based on task-specific heuristics, but by preserving key dynamical properties, allowing for efficient and generalizable representations. Beyond pruning, this dissertation extends SNNs to structured data and continuous domains through innovations such as Spiking Graph Neural Networks (SGNNs) and Spiking State-Space Models (S-SSMs), demonstrating their potential in real-world applications.Through applications spanning unsupervised learning, time-series prediction, multiagent interactions, and event-based perception, this dissertation reframes SNNs not as mere energy-efficient alternatives, but as a fundamentally new class of adaptive, real-time intelligent systems. By combining architectural innovations with deep theoretical insights, this work establishes a foundation for spike-based artificial intelligence that is not only efficient, but computationally powerful—offering a new perspective on how learning, memory, and intelligence can be reimagined through the lens of dynamical systems.

We present a Model Uncertainty-aware Differentiable ARchiTecture Search (µDARTS) that optimizes neural networks to achieve high accuracy and low uncertainty simultaneously. We introduce concrete dropout within DARTS cells and include a Monte-Carlo regularizer within the training loss to optimize the concrete dropout probabilities. A predictive variance term is introduced in the validation loss to enable searching for architecture with minimal model uncertainty. The experiments on CIFAR10, CIFAR100, SVHN, and ImageNet verify the effectiveness of µDARTS in improving accuracy and reducing uncertainty compared to existing DARTS methods. Moreover, the final architecture obtained from µDARTS shows higher robustness to noise at the input image and model parameters compared to the architecture obtained from existing DARTS methods.

Spiking Neural Networks (SNNs) represent the forefront of neuromorphic computing, promising energy-efficient and biologically plausible models for complex tasks. This paper weaves together three groundbreaking studies that revolutionize SNN performance through the introduction of heterogeneity in neuron and synapse dynamics. We explore the transformative impact of Heterogeneous Recurrent Spiking Neural Networks (HRSNNs), supported by rigorous analytical frameworks and novel pruning methods like Lyapunov Noise Pruning (LNP). Our findings reveal how heterogeneity not only enhances classification performance but also reduces spiking activity, leading to more efficient and robust networks. By bridging theoretical insights with practical applications, this comprehensive summary highlights the potential of SNNs to outperform traditional neural networks while maintaining lower computational costs. Join us on a journey through the cutting-edge advancements that pave the way for the future of intelligent, energy-efficient neural computing.

Spiking Neural Networks (SNNs) have become an essential paradigm in neuroscience and artificial intelligence, providing brain-inspired computation. Recent advances in literature have studied the network representations of deep neural networks. However, there has been little work that studies representations learned by SNNs, especially using unsupervised local learning methods like spike-timing dependent plasticity (STDP). Recent work by \\cite{barannikov2021representation} has introduced a novel method to compare topological mappings of learned representations called Representation Topology Divergence (RTD). Though useful, this method is engineered particularly for feedforward deep neural networks and cannot be used for recurrent networks like Recurrent SNNs (RSNNs). This paper introduces a novel methodology to use RTD to measure the difference between distributed representations of RSNN models with different learning methods. We propose a novel reformulation of RSNNs using feedforward autoencoder networks with skip connections to help us compute the RTD for recurrent networks. Thus, we investigate the learning capabilities of RSNN trained using STDP and the role of heterogeneity in the synaptic dynamics in learning such representations. We demonstrate that heterogeneous STDP in RSNNs yield distinct representations than their homogeneous and surrogate gradient-based supervised learning counterparts. Our results provide insights into the potential of heterogeneous SNN models, aiding the development of more efficient and biologically plausible hybrid artificial intelligence systems.

This paper shows that the heterogeneity in neuronal and synaptic dynamics reduces the spiking activity of a Recurrent Spiking Neural Network (RSNN) while improving prediction performance, enabling spike-efficient (unsupervised) learning. We analytically show that the diversity in neurons' integration/relaxation dynamics improves an RSNN's ability to learn more distinct input patterns (higher memory capacity), leading to improved classification and prediction performance. We further prove that heterogeneous Spike-Timing-Dependent-Plasticity (STDP) dynamics of synapses reduce spiking activity but preserve memory capacity. The analytical results motivate Heterogeneous RSNN design using Bayesian optimization to determine heterogeneity in neurons and synapses to improve \\(\\mathcal{E}\\), defined as the ratio of spiking activity and memory capacity. The empirical results on time series classification and prediction tasks show that optimized HRSNN increases performance and reduces spiking activity compared to a homogeneous RSNN.

Energy and data-efficient online time series prediction for predicting evolving dynamical systems are critical in several fields, especially edge AI applications that need to update continuously based on streaming data. However, current DNN-based supervised online learning models require a large amount of training data and cannot quickly adapt when the underlying system changes. Moreover, these models require continuous retraining with incoming data making them highly inefficient. To solve these issues, we present a novel Continuous Learning-based Unsupervised Recurrent Spiking Neural Network Model (CLURSNN), trained with spike timing dependent plasticity (STDP). CLURSNN makes online predictions by reconstructing the underlying dynamical system using Random Delay Embedding by measuring the membrane potential of neurons in the recurrent layer of the RSNN with the highest betweenness centrality. We also use topological data analysis to propose a novel methodology using the Wasserstein Distance between the persistence homologies of the predicted and observed time series as a loss function. We show that the proposed online time series prediction methodology outperforms state-of-the-art DNN models when predicting an evolving Lorenz63 dynamical system.

Hardware-based Malware Detectors (HMDs) have shown promise in detecting malicious workloads. However, the current HMDs focus solely on the CPU core of a System-on-Chip (SoC) and, therefore, do not exploit the full potential of the hardware telemetry. In this paper, we propose XMD, an HMD that uses an expansive set of telemetry channels extracted from the different subsystems of SoC. XMD exploits the thread-level profiling power of the CPU-core telemetry, and the global profiling power of non-core telemetry channels, to achieve significantly better detection performance than currently used Hardware Performance Counter (HPC) based detectors. We leverage the concept of manifold hypothesis to analytically prove that adding non-core telemetry channels improves the separability of the benign and malware classes, resulting in performance gains. We train and evaluate XMD using hardware telemetries collected from 723 benign applications and 1033 malware samples on a commodity Android Operating System (OS)-based mobile device. XMD improves over currently used HPC-based detectors by 32.91% for the in-distribution test data. XMD achieves the best detection performance of 86.54% with a false positive rate of 2.9%, compared to the detection rate of 80%, offered by the best performing signature-based Anti-Virus(AV) on VirusTotal, on the same set of malware samples.