Figure 9.

Histogram of the forward driving signal values using different values of b. In particular, we present the histograms for $b=1,-1,10,-10$. We include the first two since they are close to the values of b we fit from the initial 100 trials (where no deviation is applied) and the last two to illustrate the grounds on which we discard certain parameter pairs. As expected from the observed hysteresis effect (cf. Fig. 3), the histograms in (A) and (B), which correspond to $b=1$ and $b=-1$, respectively, are biased towards positive values of the driving signal. When assuming implausible parameters, like the ones in (C) and (D), which correspond to $b=10$ and $b=-10$, respectively, the bias shifts towards negative values (cf. C,D) and, even, shows a significant concentration of values around 0 (cf. C). Note we observe, respectively, the same biases in the backward driving signals.