The integer-valued parameter range used in our parameter searches depending on the neuron model [lin; sat; spk], for at most input binary variables.

In the first line, numbers give the maximal parameters values used in our searches up to . In all lines, the other numbers are the sufficient parameter values, meaning that if one was to launch a parameter search with a value higher than the one given in this table, one would find the same computation capacity as the one presented in Figure 2. For instance for , if one wants to implement all positive linearly separable Boolean function with a purely linear neuron, the first line of this table (first element between square brackets) shows that it is sufficient to have integer-valued synaptic weights between 0 and 9, and an integer-valued threshold between 0 and 18; this range is also given in [37]. We built this table using a large parameter search (see Materials and Methods), using this method we computed the Boolean function for more than parameter sets.