Silicon synapses self-correct for both mismatch and design inhomogeneities
نویسندگان
چکیده
A neuromorphic chip with an array of neurons has connections (“synapses”) which implement a biological learning mechanism known as spike-timing-dependent plasticity (“STDP”). STDP is a homeostatic mechanism which regulates the firing rate of neurons. This mechanism is shown here to reduce variation in performance between neurons, due to both mismatch in fabrication and inhomogeneities in the electronic design.
منابع مشابه
Global scaling of synaptic efficacy: Homeostasis in silicon synapses
Synaptic homeostasis is a mechanism present in biological neural systems that acts to maintain an homogeneous and stable computational substrate, in face of intrinsic inhomogeneities among neurons, and of their continuous changes due to learning processes and variations in the statistics of the input signals. In hardware spike-based neural networks homeostasis could be useful for solving issues...
متن کاملSelf-organization in Balanced State Networks by STDP and Homeostatic Plasticity
Structural inhomogeneities in synaptic efficacies have a strong impact on population response dynamics of cortical networks and are believed to play an important role in their functioning. However, little is known about how such inhomogeneities could evolve by means of synaptic plasticity. Here we present an adaptive model of a balanced neuronal network that combines two different types of plas...
متن کاملControl of bursting properties in a silicon neuron CPG
We have developed a silicon neuron based on a heart interneuron of the leech. We created a half-center oscillator composed of two silicon neurons connected via inhibitory synapses implemented through dynamic clamp. We investigated the e-ects of symmetrical variations of maximal conductances on bursting behavior. Burst period, average burst spike frequency, and duty cycle were chosen as major ch...
متن کاملDesigning Networks of Spiking Silicon Neurons and Synapses
Since the properties of transistors are well matched to the properties of ionic channels in nerve membranes, Mead has advocated the use of analog VLSI for the construction of arti cial neural systems [1]. The current state of the art in realistic single-neuron designs is the silicon neuron of Mahowald and Douglas [2], which exhibits the spiking, refractory, and adaptation characteristics of cor...
متن کاملDesign and Implementation of an Intelligent Part of Speech Generator
The aim of this paper is to report on an attempt to design and implement an intelligent system capable of generating the correct part of speech for a given sentence while the sentence is totally new to the system and not stored in any database available to the system. It follows the same steps a normal individual does to provide the correct parts of speech using a natural language processor. It...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2012