Textbook Group meeting for Parr, Pezzulo, Friston 2022 .

hello it's July 18th and we're in our first discussion of chapter three in cohort 4. so before we go to any questions or anything does anyone want to just bring up any overall thoughts they had on chapter three just about the chapter where it is in the book or what they remember or got from that chapter yeah I would add that I actually just generally think the chapter was written very well and it was very clear um I think the high road can be philosophically and mathematically very abstract so I personally appreciated the kind of clarity that the authors took to discussing things like Marco blankets and um existential imperatives I guess it's it's a the high road perhaps more than low road provokes further questions further ontological um and philosophical metaphysical questions you know the well what you know how how deep does the hierarchy of Mark of blankets run or what's the fundamental base layer of this um this whole system but on the whole at least in so far as the free energy principle is a very simple explanation of something very fundamental um yeah I thought they did a really good job outlining that particular point nice thank you anyone else any thoughts low road High Road anything yeah yeah I really enjoyed um this chapter probably because there's less equations to uh get hung up on but especially section 3.7 just the idea of using active inference to reconcile um different theoretical perspectives so in my area action we have debates about whether movement commands are represented centrally versus self-organizing or are inactive this is hotly debated for for decades now um that that's something that's really attracted me to active inference is to have more unifying view um rather than keep going with the debating so I really appreciated section 3.7 for me personally any other thoughts on three or about how somebody is seeing the low road and the high road well Daniel I want to see um about the the figures when I when I look at the figures uh for example a figure 3.1 where uh uh where the figure explains the uh blanky stays um but in addition to this Blank Space there are um many other like notations and and formulas um so it seems quite quite uh intimidating those foreigners like like uh for example the the first equation on the left you have X you know dot x equal to f of x Etc so others all this notation is not explained that's you know it seems to me um that it creates uh blocks for understanding um of the uh Concepts um so so Vision I thought it's it's kind of a little bit incompanied uh I know it's difficult because those are those in order to understand these Concepts you need more uh background uh explanations um so so but when I read those figures I I hope to understand all the information that it contains in the figures if there are things that not understandable uh that it creates a kind of a mental block uh in my mind sure totally makes sense so and it's good to want to understand all the notation so this is a representation that is going to come up again and again um it's the Markov blanket and there's kind of two uh types of notation happening here there's like the variables that are being locally assigned and then there's some more conventional notation so like some of the conventional notation is f is a function of something and then Omega is a noise term on something that's conventional and also the dot being uh rate of change so it's kind of like having like a prime or a first derivative and then there's the locally assigned variables so external X internal mu octave U sensory y and then um now there is a typo in this figure um a small one just it's in the erata okay but um the visuals are correct each of these functions each the rates of change of each of these four kinds of states we're describing the rate of change so it's kind of like a flow map for all four of these nodes on a graph for the internal external states in the blanket right rate of change yeah and each of those rates of change…