Meta-representations as representations of processes

Hello and welcome. This is ActiveInference Gueststream 79.1 on April 3rd, 2024. We're here with Ryoto Kanai talking about meta representations as representations of processes. We'll have a presentation then a discussion. So thank you very much for joining and looking forward to hearing the presentation and talking. So go for it. Yes. Okay. Right. Yeah. Yeah. Thank you very much for this opportunity to present our recent work. Yeah. And thank you for supporting our very recent paper, which just uploaded preprint. So in this talk, I want to talk about what meta representations are. And to explain my motivation, I want to start from my kind of frustration in theories of consciousness. So a lot of theories of consciousness tend to be described semantically. And then, you know, for example, like in today's talk, I mainly focus on the higher order theories of consciousness. But it kind of makes sense at the semantic level. But when we think about how we can implement such high level theories with neural networks, we realize there are a lot of uncertainties about how we might be able to implement them. So by thinking about how we might be able to implement the theories, we can actually reveal vagueness and make the concepts more precise. So that's the main topic. So as I said, I want to talk about higher order theories of consciousness. Maybe many of you might have heard of it. But just broadly speaking, a higher order theory of consciousness claims that mental state is conscious, not just by having the processing, but it becomes conscious if it's represented by a higher order mental state. So by higher order mental state or higher order representation, generally it means it's a representation about another representation. So intuitively, I think this makes sense. So for example, when your part of your brain is processing something red, people ask whether that's sufficient for consciousness, conscious experience of seeing red. But when you have an additional representation that represents that you are processing something red, so that kind of meta representation seems to be functionally related to awareness. So that kind of makes an intuitive sense. But once we start thinking about how we can implement this sort of meta representation, it's very unclear what it really means. So we try to find a way where we can sort of represent the meta representations. But let me explain why it's difficult. Okay, so this is a very naive way to think about constructing meta representation. So let's say you have an image as an input, and then there's some processing in the brain or deep neural networks, then you can represent the contents of the input or an image. But if you transform this first order representation by another neural network, that's like really naive way to implement meta representation. So let's say if we think of this as the visual information processing in the brain, maybe this could be the retina, and this could be LGN, this could be V1, V2, and so on. And but with this, somehow, like everything is just a representation of the first input. So somehow like this doesn't work. Or like if you think this is enough for meta representation, then like everything is a meta representation of something else. So yeah, so that seemed a little bit strange. And another way to conceive meta representation is to think about confidence. So but in the context of metacognition of and cognitive neuroscience, we often take confidence report from participants in an experiment. So for example, like, you know, we could present some visual stimulus, and then people report what they saw, like, you know, if it's a color discrimination task, you know, they could say blue versus red, or something like that. But then you use this first order information to make a perceptual decision. But also you can report the confidence of seeing that stimulus. So this is a frame a bit better than just the like very simplistic. So any transformation as a meta…