1
$\begingroup$

I'm reading a paper by Auge et al. (2021). In the introduction they said "Spiking Neural Networks (SNNs) use short 'all-or-nothing' pulses to encode and transmit information".

I'm not sure what is the meaning of "all-or-nothing" in this context. Does it mean binary?

Auge, D., Hille, J., Mueller, E., & Knoll, A. (2021). A survey of encoding techniques for signal processing in spiking neural networks. Neural Processing Letters, 53(6), 4693-4710.

$\endgroup$

2 Answers 2

1
$\begingroup$

For an artificial network, yes, it typically means there is a binary output; this is a standard neuron model.

For a biological network, it's a bit more complicated, because although any moment in time can be a "spike" or "no spike", there isn't some discrete timeseries described by a binary code, but rather single events with an occurrence time.

$\endgroup$
1
$\begingroup$

I guess it comes from physiology where the strength of the response of a neuron or muscle fiber doesn't have anything to do with the strength of the stimulus. If a stimulus is above a certain threshold, the neuron will fire. So, in terms of SNN, it seems to me like a binary coding, where a neuron is either active or inactive within a specific time interval.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.