# Talk:Suprathreshold stochastic resonance

Major comment: noise-aided information processing in parallel arrays of uncoupled nonlinear devices has been treated with devices presenting no threshold at all. This has been shown for example with dynamical [1] and static [2,3] systems presenting saturations with possible applications to biological neurons [2]. The interpretation of the useful-noise effect given by the authors when no threshold is at work still holds since, in fact, "suprathreshold stochastic resonance" does not require the presence of a threshold to occur. The term "suprathreshold stochastic resonance" somehow presents an historical interest due to the link with the standard stochastic resonance. A discussion about the use of the term resonance has been given by the authors. I think a discussion about the use of the term "threshold" including the following references should also be mentionned in the section dedicated to "Misconceptions about suprathreshold stochastic resonance".

[1] F. Duan, F. Chapeau-Blondeau, D. Abbott; "Input-output gain of collective response in an uncoupled parallel array of saturating dynamical subsystems"; Physica A 388, 1345-1351 (2009).

[2] S. Blanchard, D. Rousseau, F. Chapeau-Blondeau; "Noise enhancement of signal transduction by parallel arrays of nonlinear neurons with threshold and saturation"; Neurocomputing 71, 333-341 (2007).

[3] F. Chapeau-Blondeau, D. Rousseau; "Noise-aided SNR amplification by parallel arrays of sensors with saturation"; Physics Letters A 351, 231-237 (2006).

## Author McDonnell : Saturating nonlinearities

We certainly agree that suprathreshold stochastic resonance can occur for saturating nonlinearities without a "hard threshold" and have amended the article accordingly. The definition of networks in which suprathreshold stochastic resonance can occur (in the introduction) certainly allows this. Most of the focus and discussion is on binary thresholding nonlinearities because these are amenable to closed form theoretical results, and allow a stochastic quantization interpretation. But these are really special cases, and the point that "threshold" is a legacy term is quite valid.

## Reviewer B: More open research questions

The article gives a good and condensed overview of useful-noise effect in uncoupled arrays of nonlinear devices. Definitions are clear and given in a general framework which also explains to the reader the historical reasons for the name suprathreshold stochastic resonance.

I think more open research questions would be of interest in this article. The two hints given are quite application-oriented. Could the authors add some more theroretical proposals? I feel that perspectives concerning Stochastic Pooling Networks could somehow constitute such open research questions.

## Author McDonnell : More open questions

The open questions have now been categorised, and some more (very general) open theoretical questions listed. Scholarpedia articles have an encyclopedia style, so I am not sure it is appropriate to go into any more detail, as the idea is to mainly present established facts.

###### ====== Reviewer A ========

This is well written article about extremely interesting effect of positive role of noise in enhancement of information transfer in a parallel network of nonlinear elements. The authors provide a comprehensive review of work done on suprathreshold stochastic resonance. They nicely described the mechanism and discuss possible application of this phenomenon. The discussion includes clarification of misconception about stochastic resonance, which is quite useful. In my opinion, the article can be published in its present form.