 So, this is actually part of my research at the university in Amsterdam, my supervisor is Arjan van Ooyen, he's from Palt, and how he went further. So actually, before starting, I'd like you to see this video, try to identify some patterns or changes in there. Well, maybe what you can see is that during the first four bumps of activity, the inter-bombs interval is quite constant, and afterwards it changed, so it was like enlarged. So something similar happens here, for example. If you look, for example, here at the frontal zone of both humans and rats, then you can see that there is activity going on over there. If you represent here the voltage trace, and then you compute the wave electron form associated to that, representing warm colors, high activity and cold colors, low activity, you see here these two components, which are associated actually at different frequencies, but similar frequencies occur also in the rodent. So it suggests that maybe there are two different kinds of activity going on in the same region. This is synchronous activity. So what we know so far is that different networks are able to produce different kinds of synchronous activity, and it depends on the properties of the cells. If we look specifically, for example, at the prefrontal zone of the rats, layer 35 is able to oscillate at a higher frequency than layer six, but if you produce connections or you look at the connections between them, the frequencies can change. Similar dynamics can be also observed in the visual zones, comparing region, sorry, V1 and V2. But interestingly, if you look at, for example, at what happened here in this figure, this is region CA1 of the hippocampus and CA3. These regions oscillate at a higher frequency. Both are composed of excitatory inhibitor cells. The frequency here is higher than here, but by projecting connections from here, from CA3 to CA1, the frequency of the CA1 region goes down. So what it suggests is that if we actually have different rhythms, so we have two networks and we want to know how feedforward connectivity from one network is able to affect the activity in the other network. So we're going to talk from now on about two different networks. So this is our modeling approach. It was in neuron, such that we have two populations of cells. Let's say one is the source, the other is the target, but you can have one frequency lower than the other. They are non-correlated to each other. And the idea is that every network has 20 inhibitory cells and 80 excitatory cells. The cells are really abstract models, so they are conductance-based models, single compartments, including sodium, potassium, and leak currents, so that they are able to produce actually action potentials. And now that we have this structure of the cells, we connect them by using inhibitory GABA-A synapses and excitatory AMPA synapses. So what we want to know, actually, and the focus of this work is how should we connect these two networks, such that one, which is the source could be this or that, is able to impose the rhythm onto the other. So what we see is here the scheme. And to face this problem, we define here in a very systematic manner one exploration of the connectivity space, such that we define the connectivity class. And here you have the inhibitory and excitatory populations, the low-case letters for the slow network, the capital-case letters for the fast network. And then you want to modify the synaptic strength of that. But you don't have only that connection, you also have other connections there. So what we do is that we design one scheme of other connectivity situations in which we add and remove other connections without modifying the synaptic strength. So we are focusing on the strength of the main connection. So we did that by changing 10 different values, the strength of this connection, and then spanning across one range of three orders of magnitude. So it's not one linear scale. And you have in total these 32 possible schemes from the slow to the fast network. It remembers a feed-forward connection from the slow to the fast and from the fast to the slow. Now, what happens if we look at the activity in the excitatory population? Well, what you have is this picture. The inhibitory population is behaving quite similar. That's why we are just representing the excitatory one. In this situation, you have the Fourier spectrum, the wavelet transform, and here is the raster gram. So every time that one cell spikes, then you make one mark. So you can see here that this is quite stable. It's a very synchronous pattern in both situations. This is the slow network. That one is the fast network. But what happens if we connect from the slow to the fast? For example, this is an idea. So what you see is that the frequency, if you compare this row with this row, so at least by qualitatively looking first at this, you'll see one changing the frequency. The frequency is going up. But actually, if you look at the Fourier spectrum, this frequency corresponds to the first harmonic of this, when this is free. So we've made many experiments. We run 640 different cases. And three main cases is what I want to show you here. So in one case, if you connect the slow to the fast and you look at the activity at the fast, you see that in this case there is interspersed activity between the local rhythm and the external impulse rhythm. In other case, the only thing that can do is to influence the power of the oscillation in the target network, changing also the frequency. But what is interesting here is that if you, for example, connect now from the fast to the slow, and then you look at the activities in the slow network, you can observe that the frequency is not changing, but the power is quite fluctuating. Now, if we want to look at this in a systematic manner in all the simulations, what we did is that we first designed these schemes in which the red dots, which are quite small here, are representing only the location of the frequency at the base value and the first harmonic of this, in this case, the slow oscillation, which is the frequency of the source. But you can do that also for the fast. And then the blue points are representing the activity of the target network. So the size of these disks are saying how strong this oscillation is. You can do that for every connectivity class and every connectivity scheme, so you can analyze that. And at the end, you see one plethora of different kinds of patterns of activity. So in summary, what we can say is that when you get input from another network in this target, you can observe first changes in the frequency spectrum. On the other hand, you can cause the alternating occurrence of episodes of different frequencies and you can induce irregular fluctuations in the oscillation of the target network. So what we conclude is that these results suggest that the slow rather than the fast oscillation is more suitable to synchronize activity between cortical areas and combine separate representations into a whole, facilitating inter-area information transfer. Finally, I want to talk to my supervisors, my supporting project, and some colleagues at the lab. Thank you very much. So may I ask different oscillations when you talk about oscillations in C1 and 2 and 3? So can these different kinds of oscillations relate to the behaviors of these areas for hippocampus? Are they doing different kinds of memories, texts, or are they in a different phase for doing memory? Yeah, for example, in that situation, what we see is that there is synchronization between areas. What we actually want to face with this model is try to keep this property such that we observe that actually you are able to lock the activity of one network in one region to the other. We are able to reproduce this just by changing the connectivity. So that's what we want to first identify the mechanism and then look at different questions. Thank you.