Yeah actually indeed they think this is the way how it works. Now let's take a look, this is the LGN. LGN as I said is the same as the, retinal ganglion cell. This simple cell here. You have an input from three cells of the LGN. And those LGN cells or retinal ganglion cells, somehow the receptive field links to each other. [FOREIGN] So you take a look, they have some overlap here. It's response off [FOREIGN] what happened? [FOREIGN] In the center, if we have a light bar, then this cell will be very excited. And also this one, excitation, under the of this one of course excitation. If the light covered only this region, then it's inhibition of response. Also, this region is [INAUDIBLE] responsive, you can compare this site, right? [FOREIGN] Now, they also formed, during the recording, right? Of course, this is kind of called simple cell. [FOREIGN] [LAUGH] They also found some other cells. The property is actually a little bit different, take a look here. [FOREIGN] Simple cells from layers four and six mainly, right? And complex cells from two, three, five, they don't have clear excitation and inhibition regions, okay? And [FOREIGN] a bar about one third to one half of the receptive field invokes maximum responses. A stimulus [FOREIGN] receptive field [FOREIGN] cancellation, right? This actually, from the receptive field, either looks like you have really some simple interactions from the different neurons. Then you can create very dramatically different properties of the cell, right? [FOREIGN] Somehow, they'll make different combinations [FOREIGN] [FOREIGN] This is just a hypothesis, [FOREIGN]. [FOREIGN] If you have an imaging technique for this study, then you can simultaneously examine many, many different neurons. Say sometimes maybe 1,000, sometimes maybe a few hundred or a few tens of neurons, simultaneously. You just actually set the computer monitor to give the different stimulation. [FOREIGN] Of course imaging Has this kind of huge advantage at the same time may be, Not so sensitive as this kind of recording, okay, sometimes. Because if in the brain, actually those neurons actually give only a few spikes, they don't have a huge cancelling signal there. Then maybe you will miss certain information by the imaging technique. Okay, so the a lot of computations to extract those feature from the visual field, right? DM from the patch or the [FOREIGN] centers around and then to simple cell, complex cell, complex shapes. [FOREIGN] To date actually, no specific cell was found to respond only to one phase [FOREIGN] phase cell. [FOREIGN] One sale responds to the model, a phrase, that's too risky, right? [FOREIGN]