I think that's totally correct! but I'm skeptical that a lot of the non-sensory neurons work on the basis of average activity -- rate codes versus timing codes

I think that's totally correct! but I'm skeptical that a lot of the non-sensory neurons work on the basis of average activity -- rate codes versus timing codes and all of that. I spent a fair amount of time in the hippocampus where individual spikes. We were always going to run into challenges with how to interpret the voltage levels of the transistor, so we tried to be "fair" and not cheat on either side (we could have decoded the values on the memory bus or the adder numerically, for example). For the next paper, though, we're taking a more careful look at a lot of these encoding questions, as well as trying to use processors with a greater degree of functional specialization so we can more accurately capture what's going on.