arXiv subjects

Abhronil Sengupta

Publications and source records attributed to Abhronil Sengupta.

At least 37 records · Page 2Linked to original sources

arXiv2025

P-SpikeSSM: Harnessing Probabilistic Spiking State Space Models for Long-Range Dependency Tasks

Spiking neural networks (SNNs) are posited as a computationally efficient and biologically plausible alternative to conventional neural architectures, with their core computational framework primarily using the leaky integrate-and-fire (LIF) neuron model. However, the limited hidden state representation of LIF neurons, characterized by a scalar membrane potential, and sequential spike generation process, poses challenges for effectively developing scalable spiking models to address long-range dependencies in sequence learning tasks. In this study, we develop a scalable probabilistic spiking learning framework for long-range dependency tasks leveraging the fundamentals of state space models. Unlike LIF neurons that rely on the deterministic Heaviside function for a sequential process of spike generation, we introduce a SpikeSampler layer that samples spikes stochastically based on an SSM-based neuronal model while allowing parallel computations. To address non-differentiability of the spiking operation and enable effective training, we also propose a surrogate function tailored for the stochastic nature of the SpikeSampler layer. To enhance inter-neuron communication, we introduce the SpikeMixer block, which integrates spikes from neuron populations in each layer. This is followed by a ClampFuse layer, incorporating a residual connection to capture complex dependencies, enabling scalability of the model. Our models attain state-of-the-art performance among SNN models across diverse long-range dependency tasks, encompassing the Long Range Arena benchmark, permuted sequential MNIST, and the Speech Command dataset and demonstrate sparse spiking pattern highlighting its computational efficiency.

cs.NE↗

arXiv2024

Scaling SNNs Trained Using Equilibrium Propagation to Convolutional Architectures

Equilibrium Propagation (EP) is a biologically plausible local learning algorithm initially developed for convergent recurrent neural networks (RNNs), where weight updates rely solely on the connecting neuron states across two phases. The gradient calculations in EP have been shown to approximate the gradients computed by Backpropagation Through Time (BPTT) when an infinitesimally small nudge factor is used. This property makes EP a powerful candidate for training Spiking Neural Networks (SNNs), which are commonly trained by BPTT. However, in the spiking domain, previous studies on EP have been limited to architectures involving few linear layers. In this work, for the first time we provide a formulation for training convolutional spiking convergent RNNs using EP, bridging the gap between spiking and non-spiking convergent RNNs. We demonstrate that for spiking convergent RNNs, there is a mismatch in the maximum pooling and its inverse operation, leading to inaccurate gradient estimation in EP. Substituting this with average pooling resolves this issue and enables accurate gradient estimation for spiking convergent RNNs. We also highlight the memory efficiency of EP compared to BPTT. In the regime of SNNs trained by EP, our experimental results indicate state-of-the-art performance on the MNIST and FashionMNIST datasets, with test errors of 0.97% and 8.89%, respectively. These results are comparable to those of convergent RNNs and SNNs trained by BPTT. These findings underscore EP as an optimal choice for on-chip training and a biologically-plausible method for computing error gradients.

cs.NE↗

arXiv2024

Stochastic Spiking Neural Networks with First-to-Spike Coding

Spiking Neural Networks (SNNs), recognized as the third generation of neural networks, are known for their bio-plausibility and energy efficiency, especially when implemented on neuromorphic hardware. However, the majority of existing studies on SNNs have concentrated on deterministic neurons with rate coding, a method that incurs substantial computational overhead due to lengthy information integration times and fails to fully harness the brain's probabilistic inference capabilities and temporal dynamics. In this work, we explore the merger of novel computing and information encoding schemes in SNN architectures where we integrate stochastic spiking neuron models with temporal coding techniques. Through extensive benchmarking with other deterministic SNNs and rate-based coding, we investigate the tradeoffs of our proposal in terms of accuracy, inference latency, spiking sparsity, energy consumption, and robustness. Our work is the first to extend the scalability of direct training approaches of stochastic SNNs with temporal encoding to VGG architectures and beyond-MNIST datasets.

cs.NE↗

arXiv2024

Exploring Extreme Quantization in Spiking Language Models

Despite the growing prevalence of large language model (LLM) architectures, a crucial concern persists regarding their energy and power consumption, which still lags far behind the remarkable energy efficiency of the human brain. Recent strides in spiking language models (LM) and transformer architectures aim to address this concern by harnessing the spiking activity of biological neurons to enhance energy/power efficiency. Doubling down on the principles of model quantization and energy efficiency, this paper proposes the development of a novel binary/ternary (1/1.58-bit) spiking LM architecture. Achieving scalability comparable to a deep spiking LM architecture is facilitated by an efficient knowledge distillation technique, wherein knowledge from a non-spiking full-precision "teacher" model is transferred to an extremely weight quantized spiking "student" LM. Our proposed model represents a significant advancement as the first-of-its-kind 1/1.58-bit spiking LM, and its performance is rigorously evaluated on multiple text classification tasks of the GLUE benchmark.

cs.NE↗

arXiv2024

Hardware in Loop Learning with Spin Stochastic Neurons

Despite the promise of superior efficiency and scalability, real-world deployment of emerging nanoelectronic platforms for brain-inspired computing have been limited thus far, primarily because of inter-device variations and intrinsic non-idealities. In this work, we demonstrate mitigating these issues by performing learning directly on practical devices through a hardware-in-loop approach, utilizing stochastic neurons based on heavy metal/ferromagnetic spin-orbit torque heterostructures. We characterize the probabilistic switching and device-to-device variability of our fabricated devices of various sizes to showcase the effect of device dimension on the neuronal dynamics and its consequent impact on network-level performance. The efficacy of the hardware-in-loop scheme is illustrated in a deep learning scenario achieving equivalent software performance. This work paves the way for future large-scale implementations of neuromorphic hardware and realization of truly autonomous edge-intelligent devices.

cs.ET↗

arXiv2024

Deep Unsupervised Learning Using Spike-Timing-Dependent Plasticity

Spike-Timing-Dependent Plasticity (STDP) is an unsupervised learning mechanism for Spiking Neural Networks (SNNs) that has received significant attention from the neuromorphic hardware community. However, scaling such local learning techniques to deeper networks and large-scale tasks has remained elusive. In this work, we investigate a Deep-STDP framework where a rate-based convolutional network, that can be deployed in a neuromorphic setting, is trained in tandem with pseudo-labels generated by the STDP clustering process on the network outputs. We achieve $24.56\%$ higher accuracy and $3.5\times$ faster convergence speed at iso-accuracy on a 10-class subset of the Tiny ImageNet dataset in contrast to a $k$-means clustering approach.

cs.CV↗

arXiv2024

Variation-Resilient FeFET-Based In-Memory Computing Leveraging Probabilistic Deep Learning

Reliability issues stemming from device level non-idealities of non-volatile emerging technologies like ferroelectric field-effect transistors (FeFET), especially at scaled dimensions, cause substantial degradation in the accuracy of In-Memory crossbar-based AI systems. In this work, we present a variation-aware design technique to characterize the device level variations and to mitigate their impact on hardware accuracy employing a Bayesian Neural Network (BNN) approach. An effective conductance variation model is derived from the experimental measurements of cycle-to-cycle (C2C) and device-to-device (D2D) variations performed on FeFET devices fabricated using 28 nm high-$k$ metal gate technology. The variations were found to be a function of different conductance states within the given programming range, which sharply contrasts earlier efforts where a fixed variation dispersion was considered for all conductance values. Such variation characteristics formulated for three different device sizes at different read voltages were provided as prior variation information to the BNN to yield a more exact and reliable inference. Near-ideal accuracy for shallow networks (MLP5 and LeNet models) on the MNIST dataset and limited accuracy decline by $\sim$3.8-16.1% for deeper AlexNet models on CIFAR10 dataset under a wide range of variations corresponding to different device sizes and read voltages, demonstrates the efficacy of our proposed device-algorithm co-design technique.

cs.ET↗

arXiv2024

SpikingBERT: Distilling BERT to Train Spiking Language Models Using Implicit Differentiation

Large language Models (LLMs), though growing exceedingly powerful, comprises of orders of magnitude less neurons and synapses than the human brain. However, it requires significantly more power/energy to operate. In this work, we propose a novel bio-inspired spiking language model (LM) which aims to reduce the computational cost of conventional LMs by drawing motivation from the synaptic information flow in the brain. In this paper, we demonstrate a framework that leverages the average spiking rate of neurons at equilibrium to train a neuromorphic spiking LM using implicit differentiation technique, thereby overcoming the non-differentiability problem of spiking neural network (SNN) based algorithms without using any type of surrogate gradient. The steady-state convergence of the spiking neurons also allows us to design a spiking attention mechanism, which is critical in developing a scalable spiking LM. Moreover, the convergence of average spiking rate of neurons at equilibrium is utilized to develop a novel ANN-SNN knowledge distillation based technique wherein we use a pre-trained BERT model as "teacher" to train our "student" spiking architecture. While the primary architecture proposed in this paper is motivated by BERT, the technique can be potentially extended to different kinds of LLMs. Our work is the first one to demonstrate the performance of an operational spiking LM architecture on multiple different tasks in the GLUE benchmark.

cs.NE↗

arXiv2023

Hydrogen-induced tunable remanent polarization in a perovskite nickelate

Materials with field-tunable polarization are of broad interest to condensed matter sciences and solid-state device technologies. Here, using hydrogen (H) donor doping, we modify the room temperature metallic phase of a perovskite nickelate NdNiO3 into an insulating phase with both metastable dipolar polarization and space-charge polarization. We then demonstrate transient negative differential capacitance in thin film capacitors. The space-charge polarization caused by long-range movement and trapping of protons dominates when the electric field exceeds the threshold value. First-principles calculations suggest the polarization originates from the polar structure created by H doping. We find that polarization decays within ~1 second which is an interesting temporal regime for neuromorphic computing hardware design, and we implement the transient characteristics in a neural network to demonstrate unsupervised learning. These discoveries open new avenues for designing novel ferroelectric materials and electrets using light-ion doping.

cond-mat.str-el↗

arXiv2023

Sequence Learning Using Equilibrium Propagation

Equilibrium Propagation (EP) is a powerful and more bio-plausible alternative to conventional learning frameworks such as backpropagation. The effectiveness of EP stems from the fact that it relies only on local computations and requires solely one kind of computational unit during both of its training phases, thereby enabling greater applicability in domains such as bio-inspired neuromorphic computing. The dynamics of the model in EP is governed by an energy function and the internal states of the model consequently converge to a steady state following the state transition rules defined by the same. However, by definition, EP requires the input to the model (a convergent RNN) to be static in both the phases of training. Thus it is not possible to design a model for sequence classification using EP with an LSTM or GRU like architecture. In this paper, we leverage recent developments in modern hopfield networks to further understand energy based models and develop solutions for complex sequence classification tasks using EP while satisfying its convergence criteria and maintaining its theoretical similarities with recurrent backpropagation. We explore the possibility of integrating modern hopfield networks as an attention mechanism with convergent RNN models used in EP, thereby extending its applicability for the first time on two different sequence classification tasks in natural language processing viz. sentiment analysis (IMDB dataset) and natural language inference (SNLI dataset).

cs.NE↗

arXiv2023

Hybrid Stochastic Synapses Enabled by Scaled Ferroelectric Field-effect Transistors

Achieving brain-like density and performance in neuromorphic computers necessitates scaling down the size of nanodevices emulating neuro-synaptic functionalities. However, scaling nanodevices results in reduction of programming resolution and emergence of stochastic non-idealities. While prior work has mainly focused on binary transitions, in this work we leverage the stochastic switching of a three-state ferroelectric field effect transistor (FeFET) to implement a long-term and short-term 2-tier stochastic synaptic memory with a single device. Experimental measurements are performed on a scaled 28nm high-$k$ metal gate technology-based device to develop a probabilistic model of the hybrid stochastic synapse. In addition to the advantage of ultra-low programming energies afforded by scaling, our hardware-algorithm co-design analysis reveals the efficacy of the 2-tier memory in comparison to binary stochastic synapses in on-chip learning tasks -- paving the way for algorithms exploiting multi-state devices with probabilistic transitions beyond deterministic ones.

cs.ET↗

arXiv2023

Astromorphic Self-Repair of Neuromorphic Hardware Systems

While neuromorphic computing architectures based on Spiking Neural Networks (SNNs) are increasingly gaining interest as a pathway toward bio-plausible machine learning, attention is still focused on computational units like the neuron and synapse. Shifting from this neuro-synaptic perspective, this paper attempts to explore the self-repair role of glial cells, in particular, astrocytes. The work investigates stronger correlations with astrocyte computational neuroscience models to develop macro-models with a higher degree of bio-fidelity that accurately captures the dynamic behavior of the self-repair process. Hardware-software co-design analysis reveals that bio-morphic astrocytic regulation has the potential to self-repair hardware realistic faults in neuromorphic hardware systems with significantly better accuracy and repair convergence for unsupervised learning tasks on the MNIST and F-MNIST datasets. Our implementation source code and trained models are available at https://github.com/NeuroCompLab-psu/Astromorphic_Self_Repair.

cs.NE↗

arXiv2023

Leveraging Voltage-Controlled Magnetic Anisotropy to Solve Sneak Path Issues in Crossbar Arrays

In crossbar array structures, which serves as an "In-Memory" compute engine for Artificial Intelligence hardware, write sneak path problem causes undesired switching of devices that degrades network accuracy. While custom crossbar programming schemes have been proposed, device level innovations leveraging non-linear switching characteristics of the cross-point devices are still under exploration to improve the energy efficiency of the write process. In this work, a spintronic device design based on Magnetic Tunnel Junction (MTJ) exploiting the use of voltage-controlled magnetic anisotropy (VCMA) effect is proposed as a solution to the write sneak path problem. Additionally, insights are provided regarding appropriate operating voltage conditions to preserve the robustness of the magnetization trajectory during switching which is critical for proper switching probability manipulation.

cs.ET↗

arXiv2023

Leveraging Probabilistic Switching in Superparamagnets for Temporal Information Encoding in Neuromorphic Systems

Brain-inspired computing - leveraging neuroscientific principles underpinning the unparalleled efficiency of the brain in solving cognitive tasks - is emerging to be a promising pathway to solve several algorithmic and computational challenges faced by deep learning today. Nonetheless, current research in neuromorphic computing is driven by our well-developed notions of running deep learning algorithms on computing platforms that perform deterministic operations. In this article, we argue that taking a different route of performing temporal information encoding in probabilistic neuromorphic systems may help solve some of the current challenges in the field. The article considers superparamagnetic tunnel junctions as a potential pathway to enable a new generation of brain-inspired computing that combines the facets and associated advantages of two complementary insights from computational neuroscience -- how information is encoded and how computing occurs in the brain. Hardware-algorithm co-design analysis demonstrates $97.41\%$ accuracy of a state-compressed 3-layer spintronics enabled stochastic spiking network on the MNIST dataset with high spiking sparsity due to temporal information encoding.

cs.ET↗

arXiv2021

Emulation of Astrocyte Induced Neural Phase Synchrony in Spin-Orbit Torque Oscillator Neurons

Astrocytes play a central role in inducing concerted phase synchronized neural-wave patterns inside the brain. In this article, we demonstrate that injected radio-frequency signal in underlying heavy metal layer of spin-orbit torque oscillator neurons mimic the neuron phase synchronization effect realized by glial cells. Potential application of such phase coupling effects is illustrated in the context of a temporal "binding problem". We also present the design of a coupled neuron-synapse-astrocyte network enabled by compact neuromimetic devices by combining the concepts of local spike-timing dependent plasticity and astrocyte induced neural phase synchrony.

cs.ET↗

arXiv2021

Intrinsic synaptic plasticity of ferroelectric field effect transistors for online learning

Nanoelectronic devices emulating neuro-synaptic functionalities through their intrinsic physics at low operating energies is imperative toward the realization of brain-like neuromorphic computers. In this work, we leverage the non-linear voltage dependent partial polarization switching of a ferroelectric field effect transistor to mimic plasticity characteristics of biological synapses. We provide experimental measurements of the synaptic characteristics for a $28nm$ high-k metal gate technology based device and develop an experimentally calibrated device model for large-scale system performance prediction. Decoupled read-write paths, ultra-low programming energies and the possibility of arranging such devices in a cross-point architecture demonstrate the synaptic efficacy of the device. Our hardware-algorithm co-design analysis reveals that the intrinsic plasticity of the ferroelectric devices has potential to enable unsupervised local learning in edge devices with limited training data.

cs.ET↗

arXiv2021

Spin and charge interconversion in Dirac semimetal thin films

We report spin-to-charge and charge-to-spin conversion at room temperature in heterostructure devices that interface an archetypal Dirac semimetal, Cd3As2, with a metallic ferromagnet, Ni0.80Fe0.20 (permalloy). The spin-charge interconversion is detected by both spin torque ferromagnetic resonance and ferromagnetic resonance driven spin pumping. Analysis of the symmetric and anti-symmetric components of the mixing voltage in spin torque ferromagnetic resonance and the frequency and power dependence of the spin pumping signal show that the behavior of these processes is consistent with previously reported spin-charge interconversion mechanisms in heavy metals, topological insulators, and Weyl semimetals. We find that the efficiency of spin-charge interconversion in Cd3As2/permalloy bilayers can be comparable to that in heavy metals. We discuss the underlying mechanisms by comparing our results with first principles calculations.

cond-mat.mtrl-sci↗

arXiv2020

Power System Disturbance Classification with Online Event-Driven Neuromorphic Computing

Accurate online classification of disturbance events in a transmission network is an important part of wide-area monitoring. Although many conventional machine learning techniques are very successful in classifying events, they rely on extracting information from PMU data at control centers and processing them through CPU/GPUs, which are highly inefficient in terms of energy consumption. To solve this challenge without compromising accuracy, this paper presents a novel methodology based on event-driven neuromorphic computing architecture for classification of power system disturbances. A Spiking Neural Network (SNN)-based computing framework is proposed, which exploits sparsity in disturbances and promotes local event driven operation for unsupervised learning and inference from incoming data. Spatio-temporal information of PMU signals is first extracted and encoded into spike trains and classification is achieved with SNN-based supervised and unsupervised learning framework. Moreover, a QR decomposition-based selection technique is proposed to identify signals participating in the low rank subspace of multiple disturbance events. Performance of the proposed method is validated on data collected from a 16-machine, 5-area New England-New York system.

eess.SP↗