Advanced Memory and Neural Computing

This study compares spike-timing-dependent reinforcement learning and evolutionary strategy for training spiking neuronal networks to solve a reinforcement learning control task.

This paper develops a neuromorphic model of olfactory processing in Drosophila larvae, showing how sparse coding and efficient encoding can be achieved through feedback inhibition and spike frequency adaptation in a small neural network.

This paper introduces adaptive spiking recurrent neural networks with a novel surrogate gradient, achieving state-of-the-art performance in time-domain tasks and improved computational efficiency.

This paper presents a feasibility study for real-time classification of extra-cellular neural recordings using low-power VLSI systems and spike-based neuromorphic circuits.

This paper reviews spike-based symbolic computations in neural networks and shows small spiking networks can perform symbolic computations on bit sequences and numbers using biologically plausible learning rules.

Neurogrid simulates cortical models across multiple levels, revealing active dendrites mediate top-down attention.

This paper introduces a memristor-based neuromodulation device that enables real-time adaptive control of in vitro neuronal populations by detecting specific activity patterns and modulating synaptic-like interactions.

This paper proposes a neuromorphic system for real-time burst detection in neuronal cultures, achieving comparable performance to conventional methods with lower power and latency.

This paper combines spiking and non-spiking neurons to create a sensorimotor pathway that shapes network output based on analog input, validated through simulation.

CloudBrain provides an open-source cloud-based platform for neuromorphic computing that enables real-time simulation and online reconfiguration of neural networks.

This paper investigates how surrogate gradient learning performs in spiking neural networks, showing its robustness to different surrogate derivatives and the importance of regularization for effective learning.

This paper introduces a new learning method for spiking neural networks that enables biologically plausible online learning through gradient descent, achieving performance comparable to BPTT.

This paper introduces a new learning approach for spiking neural networks that enables one-shot learning through local synaptic plasticity, inspired by dopamine-based learning signals in the brain.

This paper introduces a spiking neuron and population model based on Growth Transform dynamical systems, enabling independent control of neurodynamical properties and efficient spiking associative memory.

This paper shows dendrites allow neurons to perform complex computations with less precise synaptic weights, improving efficiency in both biological and artificial systems.

This paper develops a spiking neural network that learns hand-written digits from few examples, inspired by neuroscience mechanisms, and achieves efficient learning with minimal training data.

This paper develops a learning rule for deep Spike-by-Spike networks and demonstrates its effectiveness on MNIST classification with 99.3% accuracy.

The paper develops FPGA hardware for Spike-By-Spike networks, demonstrating efficient parallelization and optimization for spiking neural network computations.

This paper introduces a framework of spiking neurons with local learning rules that achieve performance comparable to artificial deep networks in image classification.

This paper introduces a novel spiking neuron model called the Growth Transform (GT) neuron and demonstrates its application for spike-based auditory signal processing, showing stable and interpretable network dynamics.