To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/ArtemKirsanov/. The first 200 of you will get 20% off Brilliant’s annual premium subscription.
Hey Russians I have a Question Only you Can Answer: Why if Brain shows Regenerative Processes in Delta Waves during Delta wave Sleep, Telomerase that is one of the most important Regenerative Processes, if not the Stem of them all, induce a Rem Fase, Deeply Disturbed Sleeping time..? It is just Circadian Rithm ? I am Just too Sick ? [That is Most Definitely, but I Read other reports] Ergo if Delta is The Most Regenerative Status for our Brain's why Telomerase bring up the Wavelength to Something more Comparable to Rem Fase..¿ Synaptic Hyperactivity? It isn't necessarily a Bad Thing... Mullis Discovered PCR with that.. Just Saying..
Where is the evidence for "generalization"? And why would we need to generalize, when using "indirection" (as programmers use pointers) would accomplish the same thing with less work?
Artem, please consider changing the design. The pitch-black background and scorching white are hurting my eyes. I know, it’s fashionable, but for viewers with astigmatism, it’s torture to watch. I know, I belong to a minority, just spreading the word. Still, I’ll listen this video till the end.
Did you finish your degree? I tried looking you up on LinkedIn but couldn't find anything. Apologies for the intrusive question, but I promise it's for a good reason. More of a potential collaboration.
you wouldn't think of it like that, but you can feel where you are in space if you walk around your room with your eyes closed, right? thats probably the sense of the different levels those cells have.
@@xymaryai8283 I think it’s more like your brain compartmentalizes different areas of the room. In real life, your room is one continuous space, but our brain may think of it is different sections as we a walk around the room.
Yep :) I'm not sure whether "arbitrary" is a good word, though, since we don't really know how random they are. But yeah, as Ryu has mentioned in the previous comment, there are different modules of grid cells that tile the space with hexagonal patterns of different scales and phases in order to uniquely pinpoint your location
Somehow it's all drastically obvious, since you're explaining the simplest way to learn, yet simultaneously it's very fascinating to learn these basic methods.
The most curious part is where this "grid" comes from. Since many things don't need to be "relearned" does that mean these are impressions from the regularized, physics based world around us? Or is it something like a rhythm, say a cycle like the heartbeat sending signals in a relatively regular pattern that creates this "grid"?
I'm a postdoc in the intersection of neuroscience and AI, specifically about the generalization subject, and I just want to say that you are doing an amazing job communicating scientific discoveries, which is specially hard for things that happen in the brain. Huge congrats.
@@spectrumofreality long discussion. Intelligence can be defined in a variety of ways. Im tired of having this discussion, usually with people that thinks that they have all the answers already, dismissing a lot of different possibilities. *concepts do not need clarity for meaning* Think in the word "game" - an everyday word and familiar concept that we use all the time and generally don't have much trouble with, but which, under scrutiny, is very hard to define precisely You can find a variety of games... ball games, card games, games of change, games of skill, scored & unscored games, turn-based games, timed games, etc. For any explicit definition of "game" you might offer, we can find a counterexample that doesn't satisfy your definition, but which is still a game. There is no core "game-ness" concept that separates games from non-games, and there does not even appear to be any features or qualities that all games have, or that all non-games lack. This is relevant, the lack of precision does NOT mean the term "game" is meaningless, a dog is not a game, a tree is no a game, the Superbowl or the world cup final is a game, we can agree on that. The concept of "game" has fuzzy boundaries. The same is true for "understanding", "thinking", "intelligence" etc - if we define them to mean "in the exact way a human brain does”, then AIs will never qualify, but even more drastically even if you define intelligence based on several properties of humans, it is possible that you will rule out several cognitive processes of humans that don't get into your definition, but that can be considered intelligent under a slightly different one, that applies also for differences in biological species , some species do things that we can only dream with, involving several computations that we don't do, but we agree that humans are still intelligent, and that animals are intelligent. Well, moving far from agency problems, machines using AI clearly share some similarities with processes that happen in the brain, at least algorithmically, and exploring those similarities and obviously the differences, can help us to understand our own intelligence. So I never follow when a smart guy comes and say: "you are dumb, this is X and that is Y". When this debate is way more complex than that. Greetings.
This videos, apart from having concise and beautiful animation, really get to the core idea of explaining the subject. Keep going, you are doing an amazing job at neuroscience divulgation
@@waydenrd4539 was anyone on earth surprised it needed different cells for different things?? Seems like this whole video could be covered in 10seconds
Amazing video! Neuroscience for those of us without a PhD -- explained as simply as possible but no simpler. It's so fascinating to me how complicated the brain is, and yet it has these basic building block structures that we can actually make sense of.
I didnt even know it was possible to make a science video this high quality, and I say this as someone who watches them all the time! Absolutely astonishing work, I honestly wish i didn't set this video aside for so long so i could've binged your videos earlier. Thanks for making neuroscience more acessible to the public, this kind of work never gets praised enough imo
@@ryanbillie7708 no it's using that as an example. perhaps you missed the audio reference as a complement. The video is describing how our brains work. So far he has presented how we can start to think about this. One way is to say the brain uses "1-dimensional" data sets to simulate other possibilities without actually experiencing/creating memories of those possibilities.
This is really interesting when applying it to sport. It may explain "home ground advantage", and certain training strategies that could aid with spatial awareness despite sporting locations.
@Moonlight Gamers you clearly are not understanding what he meant to say... i dont know if you just want to criticise for the sake of criticising or... what he meant to say is that he likes the way he connects the broader field "neuroscience" to the minor fields "physics", "maths", etc... for example: i like the way you integrate calculus with trigonometry. I can infer from this statement that the person likes the lecturer's capability to show or explain aspects of calculus (broader) through a trigonometric (minor) lens, or aesthetic. Its not that deep.
@Moonlight Gamers yes i understand, but trust me my guy, i think youd be better off treating people with politeness and altruism, you'll find maleficent and benevolent souls in all areas of society.
Excellent editing skills, second only to your depth and exactitude. I am inspired to delve deeper into these topics and find the commonality between this and contemporary QM
Great video! I couldn't find the two references from the "Latent spaces" part so listing them here: - Hippocampal Neurons Encode Information about Different Types of Memory Episodes Occurring in the Same Location, Wood et al., 2020, - Geometry of abstract learned knowledge in the hippocampus, Nieh et al., 2021.
Hi Artem! Great video on a challenging topic. Congratulations! I like that you represented the different types of cells in the entorhinal cortex-hippocampus system as a way of a brain to label any context, whether it is qualitative relations, organization of a space or decision making. I am not a specialist in electrophysiology, but recall there were also papers on “time cells” encoding events in time. I recall my feeling from reading them - it was clear to me that it is more than simply way to encode separately different modalities in the environment but rather a way to organize it utilizing different properties. Since then, I did not follow this line of studies. And it is cool that you put the current understanding of this area together in this video. I have a critique though. To me, your example of cooking lasagna at home vs at a friend’s kitchen is somewhat confusing as an illustration of brain’s ability to generalize (adopt). The fact is that if we want to illustrate the generalization of a cooking skill, then we must show that a person can cook lasagna from different ingredients. Whereas if we want to illustrate the generalization of a skill to navigate in a different space of a friend’s kitchen (with a complex mix of similar and dissimilar properties) then (as in your example) we may show that a person can recognize and discriminate tables, stoves and refrigerators. In your example, everything is a little bit mixed together. Anyway, consider this as a minor revision! Overall, a solid and inspiring video! Cheers!
14:01 “To completely capture all the relevant information about this task you need a configuration of a cognitive map which keeps track of both the location in the physical 2D space of the maze and a binary location in this abstract space of left and right trials.” No, you don’t. All the rat has to do-and probably does-is behave _with respect to the space,_ specifically the walls of the maze and itself. When it reaches the point of the T where it can no longer move forward, to get the reward it turns one way or the other. It turns the opposite way with respect to itself from that in which it turned in the previous trial. That’s a far more parsimonious and likely explanation. There’s no “cognitive map” keeping track of “the location in the physical 2D space of the maze” or “binary location in this abstract space” required.
ปีที่แล้ว +20
Excellent content! I would be very curious to see how this research is conducted, like how do you record neural activity of a mouse brain while it's moving through a maze? Maybe you can make a video about your lab at some point?
Thank you! Totally, I was actually considering making a video like this for quite a while :) I agree, it would be more engaging to film a video like this in a real lab. Right now I'm finishing my bachelor's degree (which is mostly computational) and transitioning to begin experimental work in graduate school this fall. So, hopefully, at some point, I would be able talk about the experimental side of things first-hand 🐁
Fascinating video! Next year I am going to college to study cognitive science. So as a request for future videos, I would appreciate it if you could further explore the interdisciplinary nature of neuroscience because it would nicely complement what I learn from my studies. Thanks
Yet another outstanding video. Thank you for making such high quality content. I look forward to the second part discussing the Tolman Eichenbaum Machine. It is especially interesting, considering they were able to replicate the results by modifying a transformer, the backbone architecture to popular LLMs such as ChatGPT. Again thanks for providing the highest quality computational neuroscience content on youtube
Thank you so much for all the effort to make these videos, I found your channel yesterday and is amazing, it's one of the best I've seen on TH-cam, especially this video. I loved the way you explain all these abstract mathematical concepts in such a clear way. Greetings from Chile!
this is incredible, now that i'm aware of the system you mentioned, about the hexagonal points in space mammals seem to use to find their way, i can *feel* them as i move around.
I've watched this at least a couple of times now. Is an astoundingly well put together video. Easy to understand despite being quite advanced concepts that would usually take much longer to put together. You really have a talent for translating insightful yet advanced concepts into language anyone can understand. By the way, I assume others have pointed this out, but your "east" arrow was pointing West 😂 Honestly, one of my top 3 channels on TH-cam. All the hard work is much appreciated!
Impressive quality! Can't wait for part 2. If you've heard about Vicarious AI (which is trying to solve some of the things talked in the video) would love to hear your thoughts about it.
Overgeneralization is where most Mandela effects come from. When remembering things we vaguely remember something and when our brain proves puts it together it pulls 2 or more similar but different things together into one memory that you believe.
Really interesting, thanks for your work! But I feel you didn't mention an important point; quid of the mapping between the environment and the enthorinal cortex? If you have a small environment or a big environment, does it mean the hexagonal pattern becomes sparse, relative or just repeats? What if you stretch or contract the environment? Then quid of the unbounded walls (like the horizon)? How do you properly map such a situation ad infinitum? And I'm also wondering; what about those directed mapping relatives to objects? Are those objects maps then a direction is added? Are all directions consider (which would be an ad infinitum problem as well), only a few of those or only one? If only one is considered, is it determined relative to us or relative to it? And would that be why we have this tendency to put a front and a back to everything? What about a sphere? => As I can picture that there is a part of the sphere facing me and as I can consider that people have a front anyway if they're facing me or are not part of any scene (like people represented in my mind independently of what they're facing; I would guess that either we have both direction existing or we can switch to both and integrate the information of where's the front and what is it facing in a more complete representation)
This is what I'm talking about! It's very rare to find videos that talk about the mechanisms of function in the brain - they're almost always purely empirical and focus on biology (i.e. this part of the brain is responsible for X etc), or very vague and broad without really suggesting any specific mechanisms of cognition. Keep em coming & thanks for posting! Btw I would love to know your thoughts on H.M. and how he remained cognitively normal despite destruction of hipp, and the anterograde amnesia. And perhaps the mechanisms of voluntary control in relation to memory formation and lookup.
Hey, that's stuff I've been working on for the past couple years, although in relation to artificial neural nets! So nice to see it works the same way with brains, haha! I am asking myself all the time how much knowledge is transferable between neuroscience and transformer-based networks, because it seems like A LOT. I guess knowing both is like being bilingual. This is why, if you haven't yet, you may be interested in "Topology of Deep Neural Networks". Also, you should try interrogating GPT-4, it interpolates between the two disciplines very well and it's a lot of fun!
Thank you! I've just read the abstract for the "Topology of Deep Neural Networks" and it sounds like an incredibly promising approach! Can't wait to read the full paper! When I first saw the title I thought it was about the topology of the networks themselves (in terms of connectivity between the layers), rather than how the _topology of the data_ changes as it flows though the network. So personally I would have named the paper a bit differently, but that's just nitpicking :D Perhaps you would be interested to know, that the Tolman-Eichenbaum Machine - a computational model of the hippocampus and the topic of the next video - is actually a modification of a *Transformer network* ! I'm not sure whether I'm going to dive deep in this equivalence in the upcoming video, but here's the brilliant paper about relating Transformers to hippocampus, where they show that when the model is trained to navigate on graphs, grid, place and boundary cells emerge naturally: arxiv.org/abs/2112.04035
@@ArtemKirsanov A few days ago, I asked Claude+ "What does physically happen in your brain, when after having exerted a prolonged mental effort you feel unable to focus?" and look at points 4-7. They look very interesting in light of several of your videos. Claude+: Several things are likely happening in the brain when mental fatigue sets in after prolonged mental effort: 1. Depletion of mental resources: The brain requires glucose and oxygen to function properly. After prolonged mental activity, the brain's stores of glucose and oxygen can get depleted, leading to fatigue. 2. Buildup of waste products: As the brain is active, waste products such as lactate accumulate. Too much buildup of these waste products can impair brain function and lead to fatigue. 3. Neurotransmitter imbalance: Prolonged activity can lead to depletion or imbalance of neurotransmitters such as dopamine that are important for attention, focus and motivation. This can contribute to fatigue. 4. Decreased activation of attention networks: Parts of the brain involved in attention and focus, especially the dorsolateral prefrontal cortex, show reduced activation after prolonged mental activity. This makes it harder to sustain attention and concentrate. 5. Increased mind wandering: When mentally fatigued, there is more erratic activity across multiple brain networks, indicating increased mind wandering and distractibility. This makes it difficult to focus. 6. Altered brain wave patterns: There is a shift in the brain's electrical activity or brain waves, away from more alert patterns to slower waves as fatigue sets in. Slower, more irregular waves are associated with lower arousal and alertness. 7. Disrupted neural connectivity: Temporary mental fatigue may alter the levels of connectivity between different parts of the brain, disrupting optimal communication and coordination between brain networks. This impacts attention, focus and other functions. With rest, all of these changes are reversible as the brain restores its mental and metabolic resources. Getting adequate sleep, taking breaks, and reducing cognitive load can all help prevent and relieve mental fatigue.
@@matveyshishov or you work through that mental fatigue and wait till your live does its damn glycolysis work. Nah kidding this is great because I've been forcing myself to focus on this video right now because mental fatigue did set in. But i also thing that it's attention span based, lien you say wandering off, i wander off seeking the next thing that would release more dopamine than previous. So yeah it's depletion of Neurotransmitter but it's also based on learned imbalance, meaning other shit is more tempting and will release the dopamine needed to keep going isntead of focusing and having that delayed more longer reward pathway going. This is why I focused on that video, and reading and writing this comment give me the closure i need for my brain to be more sure through a successful task that it was worth the forced focus amking more apt to have higher attention spans in the future. Do thanks for providing me that actionist with your comment. Haha
I believe that it would be more correct with regards to factorization that the categories are: (1) object qualities and (2) valuation/dynamics, not "sensory". All organisms use their senses to discern features or qualities, but then must relate these to valuation and dynamics within a context. The term sensory as used in this video would be imprecise within the discussion of factorization. How one "feels" about something is grounded on the recognized valuation within a context.
So the brain has learned to recognize the repetitive nature of spatial and non spatial variables in nature and in order to become as efficient as possible at processing information and solving problems critical for survival, it evolved the ability of factorization of different experiences and also evolved specialized cells to encode those variables in a "mental graph" and the ability to then combine different graphs into a higher dimensional graph that is specialized at solving the particular problem the brain is presented at any given circumstance. At least that's what I understand as a layman. Great video as always. I really like your content and although I'm not a neuroscientist myself I find the field more and more fascinating.
You, my friend, should be regarded as an international treasure. Im speechless at how great this video is in all facets i can think of right now. Thank you!
This video is amazingly clean and beautifully put together/choerent. I am appalled this video (and the rest of your videos) don't have millions of views (yet). Even if I have barely any hope of understanding most of what was said in this video, you came as close as possible to making me I think I do. And either way, if this entire video was in a completely foreign language with no subtitles, I would still have watched the whole thing for the visuals alone. Fantastic
Very nicely explained and interresting, just a little detail that I guess nobody noticed.. or was it on purpose to demonstrate what the brain is capable of.... at 3:45.. "neural represenations"... I have subscribed and eager to learn more, thank you very much for this good work of sharing knowledge to everybody!!
Hi Artem, thank you for your fantastic videos! I really enjoy them and admire your ability to explain visualize such topic so elegantly and comprehensive! Brilliant visuals! Small glitch at around 10:43 path integration: East arrow points West.
The only time TH-cam adds value to my life is when I'm watching Artem. The rest is attention and time theft. Thank you 🙏🏾 for the premium quality content.
🎯 Key Takeaways for quick navigation: 00:00 🧠 El cerebro puede generar comportamientos flexibles generalizados en diferentes contextos, gracias a su capacidad de generalización. 02:13 🐀 En experimentos con ratas, se observa que tienen la capacidad de utilizar un mapa cognitivo interno para tomar decisiones en laberintos, incluso en situaciones nuevas. 04:06 🧭 La formación del hipocampo y la corteza entorrinal son cruciales para la construcción de mapas cognitivos en el cerebro. 06:07 🧭 Los tipos de neuronas, como las células de lugar y las células de cuadrícula, juegan un papel importante en la creación de mapas cognitivos. 11:13 🧲 Los mapas cognitivos se asemejan a estructuras de grafo que representan relaciones entre elementos del mundo. 14:53 🗺️ Los mapas cognitivos pueden expandirse para incluir información abstracta, como direcciones de giro en un laberinto. 17:08 🔗 La factorización de conocimiento en componentes estructurales y sensoriales permite al cerebro generalizar y simplificar la resolución de problemas. 19:57 🧩 Las células en el hipocampo y la corteza entorrinal ayudan a factorizar y conjugar información estructural y sensorial en los mapas cognitivos. 23:13 🗺️ Los mapas cognitivos organizan el conocimiento de manera eficiente, permitiendo la generalización y resolución de problemas en diferentes contextos. Made with HARPA AI
🎯 Key Takeaways for quick navigation: 00:00 🧠 Introduction to Cognitive Mapping - The brain's ability to generate flexible behavior and generalize across contexts. - Generalization explained through the concept of cognitive maps. - Introduction to the two-part series on cognitive mapping. 01:39 🗺️ Cognitive Maps and Spatial Behavior - Overview of the Tolman agenda machine and its role in solving the problem of cognitive mapping. - Edward Tolman's experiment with rats in different mazes. - Historical context of cognitive maps in spatial behavior. 03:53 🧭 Hippocampal Formation and Spatially Selective Neurons - The role of the hippocampal formation in cognitive mapping. - Introduction to place cells, grid cells, object vector cells, and landmark cells. - Highlighting the diversity of cellular responses in the hippocampal formation. 06:07 🧩 Organization of Cognitive Maps - Explanation of how entorhinal cortex and hippocampus work together. - The general coordinate system provided by entorhinal cortex and the specific code from the hippocampus. - Examples of spatial and non-spatial tasks influencing hippocampal neurons. 08:44 📊 Graph Theory and Cognitive Maps - Introduction to graph theory as a mathematical formalism for organizing information. - The importance of knowing one's location on the graph for effective cognitive maps. - Path integration as a mechanism for updating one's position on the graph. 11:55 🌐 Unifying Spatial and Non-Spatial Domains - Conceptual unification of spatial and non-spatial tasks using graph theory. - The versatility of the brain in solving computational tasks across different modalities. - The role of relational graphs in the brain's cognitive processes. 17:25 🧠 Latent Spaces and Factorized Representations - Introduction to latent spaces and their importance in cognitive maps. - The role of latent spaces in capturing abstract variables like sensory evidence. - The concept of factorized representations for computational efficiency. 19:57 🤖 Generalization and Factorization in Cognitive Maps - How factorization allows the brain to generalize knowledge. - Comparing factorized and conjunctive representations in the hippocampal formation. - Evidence for factorization in the responses of individual neurons. 22:05 🌐 Summary and Cognitive Mapping Applications - Recap of behavioral and cellular evidence for cognitive maps. - The role of cognitive maps in organizing knowledge efficiently. - Teaser for the second part of the series focusing on building a computational system. Made with HARPA AI
11:29 I know it's beside the point, but it's worth noting that human families, species taxonomy, and language taxonomy all differ from that idealized model in different ways when you try to understand them more accurately.
It's videos like this that support some personal theories about the world and natural laws, that touch on other theory's, and then leave me with a better respect for what it all is. Humbling myself in the pursuit of better understanding. Thank you.
I have to say the production of your videos is superb (I can't think of another word). Well structured and animated. The only presentation tip I could give you, is to use short pauses on the conclusion and on important notes or exciting facts. Otherwise perfect!
What are your thoughts on the Forward Model mechanism for planning and coordination? I'm wondering if you can also make a video explaining the relationship between cognitive map and FM. Thanks for making such amazing content!
the title of this video really made me think you were gonna talk about the recent Tim Berhrens paper on HPC and mPFC representing the specifics and the generals of tasks… but nonetheless, nice video.
Thanks for presenting such a cool topic in such an accessible way! Your clear explanations and engaging visuals really made a complex subject easy to grasp. Fantastic work!
Thank you so much for making this video! It helped me finally understand what makes me so interested in neuroscience, AI - the principles that let things organize themselves. Great job, I’ll have to do some further reading!
Wow, what an amazing video, very well explained and simplified just in the right amount so that we can understand it but we know there's a lot more about the subject to be studied. Great job man, these videos breaking down how we learn at a fundamental level are really helpful for creating better studying techniques. Thank you and keep it up!
I'm very late to the party I see, but I'd just like to congratulate you on an excellent educational AND entertaining video; I'm highly impressed! Not only do you 1. use simple but beautiful and engaging animations, 2. you have a "story to tell" from A to Z (well ok there's a part 2 but I'm going there next), 3. you use great analogies, 4. explain jargon/terminology, you 5. summarize in between and at the end (spaced repetition!) and 6. your voice adds to the visuals instead of simply copying. 6 out of 5 stars from me, and an immediate subscribe! Best of luck to you, you really have a talent for education! (See, i factorized my critique, proving i actually learned something ;-)
Coming from a computer science background I'm relating all this to a CPU. The hippocampus is like an ALU, dedicated hardware to accelerate certain tasks. Is there a structure in the brain analogous to the control logic? Something that directs all this "hardware" to perform the task
Two things that came to my mind. 1. This video reminds me of when I studied Kant's Aesthetics and philosophy in general during high school. 2. Now I wonder what goes on in a mathematician's mind when he's writing a proof to a theorem, or a surgeon applying his knowledge to a case he's going to operate.
Hexagonal grid: Hmm, is it really hexagonal, or simply triangular, as depicted in the graphics. Obviously a triangular grid contains hexagons, but the difference is that a hexagonal grid is missing some nodes (the centers of the hexagons) compared to a triangular grid.
I had a weird experience a couple of years ago where I was out for a walk and I suddenly felt like I was on a street which I had completely forgotten about for 7 years, in a different country. Could that have been caused by the relative placement of houses and roads suddenly triggering some place cells and grid cells in such a way that brought that all back?
your logo is oroboros/infinity inside the frame of a human brain,yes? fantastic video! so much new details i gad no idea about. thanks for sharing,man!
I can't wait to go through all your videos. Nice intro to these subjects. The one thing that I keep admiring is the visualisations. Are you producing these things alone or with help? How long does it take to create and put together all the visualisations for the video?
![The moment we stopped understanding AI [AlexNet]](/img/n.gif)
To try everything Brilliant has to offer-free-for a full 30 days, visit brilliant.org/ArtemKirsanov/.
The first 200 of you will get 20% off Brilliant’s annual premium subscription.
Hey Russians I have a Question Only you Can Answer:
Why if Brain shows Regenerative Processes in Delta Waves during Delta wave Sleep, Telomerase that is one of the most important Regenerative Processes, if not the Stem of them all, induce a Rem Fase, Deeply Disturbed Sleeping time..?
It is just Circadian Rithm ?
I am Just too Sick ?
[That is Most Definitely, but I Read other reports]
Ergo if Delta is The Most Regenerative Status for our Brain's why Telomerase bring up the Wavelength to Something more Comparable to Rem Fase..¿
Synaptic Hyperactivity?
It isn't necessarily a Bad Thing...
Mullis Discovered PCR with that..
Just Saying..
Where is the evidence for "generalization"? And why would we need to generalize, when using "indirection" (as programmers use pointers) would accomplish the same thing with less work?
Artem, please consider changing the design. The pitch-black background and scorching white are hurting my eyes. I know, it’s fashionable, but for viewers with astigmatism, it’s torture to watch. I know, I belong to a minority, just spreading the word. Still, I’ll listen this video till the end.
Did you finish your degree? I tried looking you up on LinkedIn but couldn't find anything. Apologies for the intrusive question, but I promise it's for a good reason. More of a potential collaboration.
structure and location are different? everybody knows that? can you explain the point of over 25 minutes talking about what it is like to walk around?
Bold of you to assume my brain organises information. My brain organises information like a mailbox leading into a fucking shredder
Be nice to yourself 😊
wait are you telling me there is an arbitrary hexagonal grid of points in my room where specific cells light up the most?
you wouldn't think of it like that, but you can feel where you are in space if you walk around your room with your eyes closed, right? thats probably the sense of the different levels those cells have.
@@xymaryai8283 I think it’s more like your brain compartmentalizes different areas of the room. In real life, your room is one continuous space, but our brain may think of it is different sections as we a
walk around the room.
It's actually many grids with different scales, rotations, and offsets, but yeah.
Yep :)
I'm not sure whether "arbitrary" is a good word, though, since we don't really know how random they are.
But yeah, as Ryu has mentioned in the previous comment, there are different modules of grid cells that tile the space with hexagonal patterns of different scales and phases in order to uniquely pinpoint your location
Replace Rat 🐀 with Pigeon 🐦. Some Beehive 🐝 and Ant Colonies 🐜 visuals can help in study and understanding the abstraction concepts.
Somehow it's all drastically obvious, since you're explaining the simplest way to learn, yet simultaneously it's very fascinating to learn these basic methods.
The most curious part is where this "grid" comes from. Since many things don't need to be "relearned" does that mean these are impressions from the regularized, physics based world around us? Or is it something like a rhythm, say a cycle like the heartbeat sending signals in a relatively regular pattern that creates this "grid"?
@@CarlosVixil Surely any periodicity, external or consequent to the system itself, can be used as a grid -- no need to pick a single source.
I'm a postdoc in the intersection of neuroscience and AI, specifically about the generalization subject, and I just want to say that you are doing an amazing job communicating scientific discoveries, which is specially hard for things that happen in the brain. Huge congrats.
Neuroscience MSc here - couldn't agree more! This is excellent stuff.
There's no such thing as A.I. it's a misnomer to describe algorythms and machine learning which is nowhere in the realms of intelligence...
@@spectrumofreality long discussion. Intelligence can be defined in a variety of ways.
Im tired of having this discussion, usually with people that thinks that they have all the answers already, dismissing a lot of different possibilities.
*concepts do not need clarity for meaning*
Think in the word "game" - an everyday word and familiar concept that we use all the time and generally don't have much trouble with, but which, under scrutiny, is very hard to define precisely
You can find a variety of games... ball games, card games, games of change, games of skill, scored & unscored games, turn-based games, timed games, etc.
For any explicit definition of "game" you might offer, we can find a counterexample that doesn't satisfy your definition, but which is still a game.
There is no core "game-ness" concept that separates games from non-games, and there does not even appear to be any features or qualities that all games have, or that all non-games lack.
This is relevant, the lack of precision does NOT mean the term "game" is meaningless, a dog is not a game, a tree is no a game, the Superbowl or the world cup final is a game, we can agree on that.
The concept of "game" has fuzzy boundaries.
The same is true for "understanding", "thinking", "intelligence" etc - if we define them to mean "in the exact way a human brain does”, then AIs will never qualify, but even more drastically even if you define intelligence based on several properties of humans, it is possible that you will rule out several cognitive processes of humans that don't get into your definition, but that can be considered intelligent under a slightly different one, that applies also for differences in biological species , some species do things that we can only dream with, involving several computations that we don't do, but we agree that humans are still intelligent, and that animals are intelligent.
Well, moving far from agency problems, machines using AI clearly share some similarities with processes that happen in the brain, at least algorithmically, and exploring those similarities and obviously the differences, can help us to understand our own intelligence.
So I never follow when a smart guy comes and say: "you are dumb, this is X and that is Y". When this debate is way more complex than that.
Greetings.
@nunezkant9835 what a lovely simile. Never heard it put quite like that.
@@spectrumofreality You must be fun at parties. And what are you, 14? No way you just said that to a literal doctor, thinking you're being smart.
Thanks!
Wow, thank you!!
This videos, apart from having concise and beautiful animation, really get to the core idea of explaining the subject. Keep going, you are doing an amazing job at neuroscience divulgation
Thank you!
core of the idea of walking around...
@@ryanbillie7708 what do you mean ?
@@waydenrd4539 was anyone on earth surprised it needed different cells for different things?? Seems like this whole video could be covered in 10seconds
Amazing video! Neuroscience for those of us without a PhD -- explained as simply as possible but no simpler. It's so fascinating to me how complicated the brain is, and yet it has these basic building block structures that we can actually make sense of.
Thank you! I'm really glad to hear that!
@@ArtemKirsanov do you think these concepts this could be applicable to the design of AGI systems?
You don't have to be a neuroscience major to know this lol
@@petricor1420 fascinating and useful comment
@@lunafoxfire it’s ok, he even spelled his own username wrong
I didnt even know it was possible to make a science video this high quality, and I say this as someone who watches them all the time! Absolutely astonishing work, I honestly wish i didn't set this video aside for so long so i could've binged your videos earlier. Thanks for making neuroscience more acessible to the public, this kind of work never gets praised enough imo
I've been teaching this subject matter for three years, and you've done a great job of simplifying the subject.
is this video really just talking about walk it is like to walk around?
@@ryanbillie7708 no it's using that as an example. perhaps you missed the audio reference as a complement. The video is describing how our brains work. So far he has presented how we can start to think about this. One way is to say the brain uses "1-dimensional" data sets to simulate other possibilities without actually experiencing/creating memories of those possibilities.
@@CarlosVixil But... but walk...
At 10:30, when describing the positions, both arrows are pointing in the same direction:
Amazing video BTW, THANK YOU!! PLEASE KEEP DOING THIS AMAZING JOB!
This is really interesting when applying it to sport. It may explain "home ground advantage", and certain training strategies that could aid with spatial awareness despite sporting locations.
This content is so good I draw mind maps and summarise on a massive whiteboard while I listen to you. It’s great content, thank you.
This is mental....amazing work!
10:37 i believe the intention was to have also E and W going in opposite directions 😂
Super informative videos. Good work.
So hexagons really are the bestagons
AI/HPC researcher here.
I just love how you integrate neuroscience, math, physics and computing in your stubs ❤
Greetings from Brazil
hu3 hu3 teach metal brain how to samba.
@Moonlight Gamers he clearly said "HOW" he integrates not "THAT" he integrates.
@Moonlight Gamers you clearly are not understanding what he meant to say... i dont know if you just want to criticise for the sake of criticising or... what he meant to say is that he likes the way he connects the broader field "neuroscience" to the minor fields "physics", "maths", etc... for example: i like the way you integrate calculus with trigonometry. I can infer from this statement that the person likes the lecturer's capability to show or explain aspects of calculus (broader) through a trigonometric (minor) lens, or aesthetic. Its not that deep.
@Moonlight Gamers yes i understand, but trust me my guy, i think youd be better off treating people with politeness and altruism, you'll find maleficent and benevolent souls in all areas of society.
That opening line itself made me subscribe. Flexible behavior generalized across contexts. Chef's kiss.
good video! Thanks share
thank you for taking your viewer's suggestions into account! I especially appreciated the explanation of factorization
Artem, mate, your videos are just amazing. You are the future of education. Keep up the stunning work!
3:14 -- Amazing transition: visuals + voice + story flow work perfectly!
Excellent editing skills, second only to your depth and exactitude. I am inspired to delve deeper into these topics and find the commonality between this and contemporary QM
Impressive! I’m a M1 student in Neuroscience who’s interested in the computational aspect. Your contents are very helpful !
Your visual and verbal explanatory ability are absolutely masterful.
Great video! I couldn't find the two references from the "Latent spaces" part so listing them here:
- Hippocampal Neurons Encode Information about Different Types of Memory Episodes Occurring in the Same Location, Wood et al., 2020,
- Geometry of abstract learned knowledge in the hippocampus, Nieh et al., 2021.
Hi Artem! Great video on a challenging topic. Congratulations! I like that you represented the different types of cells in the entorhinal cortex-hippocampus system as a way of a brain to label any context, whether it is qualitative relations, organization of a space or decision making. I am not a specialist in electrophysiology, but recall there were also papers on “time cells” encoding events in time. I recall my feeling from reading them - it was clear to me that it is more than simply way to encode separately different modalities in the environment but rather a way to organize it utilizing different properties. Since then, I did not follow this line of studies. And it is cool that you put the current understanding of this area together in this video.
I have a critique though. To me, your example of cooking lasagna at home vs at a friend’s kitchen is somewhat confusing as an illustration of brain’s ability to generalize (adopt). The fact is that if we want to illustrate the generalization of a cooking skill, then we must show that a person can cook lasagna from different ingredients. Whereas if we want to illustrate the generalization of a skill to navigate in a different space of a friend’s kitchen (with a complex mix of similar and dissimilar properties) then (as in your example) we may show that a person can recognize and discriminate tables, stoves and refrigerators. In your example, everything is a little bit mixed together.
Anyway, consider this as a minor revision! Overall, a solid and inspiring video!
Cheers!
14:01 “To completely capture all the relevant information about this task you need a configuration of a cognitive map which keeps track of both the location in the physical 2D space of the maze and a binary location in this abstract space of left and right trials.”
No, you don’t. All the rat has to do-and probably does-is behave _with respect to the space,_ specifically the walls of the maze and itself. When it reaches the point of the T where it can no longer move forward, to get the reward it turns one way or the other. It turns the opposite way with respect to itself from that in which it turned in the previous trial. That’s a far more parsimonious and likely explanation. There’s no “cognitive map” keeping track of “the location in the physical 2D space of the maze” or “binary location in this abstract space” required.
Excellent content! I would be very curious to see how this research is conducted, like how do you record neural activity of a mouse brain while it's moving through a maze? Maybe you can make a video about your lab at some point?
Thank you! Totally, I was actually considering making a video like this for quite a while :)
I agree, it would be more engaging to film a video like this in a real lab. Right now I'm finishing my bachelor's degree (which is mostly computational) and transitioning to begin experimental work in graduate school this fall.
So, hopefully, at some point, I would be able talk about the experimental side of things first-hand 🐁
Fascinating video! Next year I am going to college to study cognitive science. So as a request for future videos, I would appreciate it if you could further explore the interdisciplinary nature of neuroscience because it would nicely complement what I learn from my studies. Thanks
Yet another outstanding video. Thank you for making such high quality content.
I look forward to the second part discussing the Tolman Eichenbaum Machine. It is especially interesting, considering they were able to replicate the results by modifying a transformer, the backbone architecture to popular LLMs such as ChatGPT.
Again thanks for providing the highest quality computational neuroscience content on youtube
The production quality on this video is incredible. Amazing work
Thank you so much for all the effort to make these videos, I found your channel yesterday and is amazing, it's one of the best I've seen on TH-cam, especially this video. I loved the way you explain all these abstract mathematical concepts in such a clear way. Greetings from Chile!
this is incredible, now that i'm aware of the system you mentioned, about the hexagonal points in space mammals seem to use to find their way, i can *feel* them as i move around.
Just commenting to say I thoroughly enjoyed this video. More power to you.
Quality of content really improving together with accuracy of animation, making videos more interesting and clear, very good job man!
Thanks!
I've watched this at least a couple of times now. Is an astoundingly well put together video. Easy to understand despite being quite advanced concepts that would usually take much longer to put together.
You really have a talent for translating insightful yet advanced concepts into language anyone can understand.
By the way, I assume others have pointed this out, but your "east" arrow was pointing West 😂
Honestly, one of my top 3 channels on TH-cam. All the hard work is much appreciated!
The visual and explanatory parts are insanely good! ✨
Thank you for the quality content 😊👍
Impressive quality! Can't wait for part 2. If you've heard about Vicarious AI (which is trying to solve some of the things talked in the video) would love to hear your thoughts about it.
Overgeneralization is where most Mandela effects come from. When remembering things we vaguely remember something and when our brain proves puts it together it pulls 2 or more similar but different things together into one memory that you believe.
Really interesting, thanks for your work!
But I feel you didn't mention an important point; quid of the mapping between the environment and the enthorinal cortex?
If you have a small environment or a big environment, does it mean the hexagonal pattern becomes sparse, relative or just repeats? What if you stretch or contract the environment?
Then quid of the unbounded walls (like the horizon)? How do you properly map such a situation ad infinitum?
And I'm also wondering; what about those directed mapping relatives to objects? Are those objects maps then a direction is added? Are all directions consider (which would be an ad infinitum problem as well), only a few of those or only one? If only one is considered, is it determined relative to us or relative to it? And would that be why we have this tendency to put a front and a back to everything? What about a sphere?
=> As I can picture that there is a part of the sphere facing me and as I can consider that people have a front anyway if they're facing me or are not part of any scene (like people represented in my mind independently of what they're facing; I would guess that either we have both direction existing or we can switch to both and integrate the information of where's the front and what is it facing in a more complete representation)
This is what I'm talking about! It's very rare to find videos that talk about the mechanisms of function in the brain - they're almost always purely empirical and focus on biology (i.e. this part of the brain is responsible for X etc), or very vague and broad without really suggesting any specific mechanisms of cognition.
Keep em coming & thanks for posting!
Btw I would love to know your thoughts on H.M. and how he remained cognitively normal despite destruction of hipp, and the anterograde amnesia. And perhaps the mechanisms of voluntary control in relation to memory formation and lookup.
ngl super super super underrated channel... i never noticed before... keep up the insanely high quality work
Hey, that's stuff I've been working on for the past couple years, although in relation to artificial neural nets! So nice to see it works the same way with brains, haha!
I am asking myself all the time how much knowledge is transferable between neuroscience and transformer-based networks, because it seems like A LOT. I guess knowing both is like being bilingual.
This is why, if you haven't yet, you may be interested in "Topology of Deep Neural Networks".
Also, you should try interrogating GPT-4, it interpolates between the two disciplines very well and it's a lot of fun!
Thank you! I've just read the abstract for the "Topology of Deep Neural Networks" and it sounds like an incredibly promising approach! Can't wait to read the full paper!
When I first saw the title I thought it was about the topology of the networks themselves (in terms of connectivity between the layers), rather than how the _topology of the data_ changes as it flows though the network. So personally I would have named the paper a bit differently, but that's just nitpicking :D
Perhaps you would be interested to know, that the Tolman-Eichenbaum Machine - a computational model of the hippocampus and the topic of the next video - is actually a modification of a *Transformer network* ! I'm not sure whether I'm going to dive deep in this equivalence in the upcoming video, but here's the brilliant paper about relating Transformers to hippocampus, where they show that when the model is trained to navigate on graphs, grid, place and boundary cells emerge naturally: arxiv.org/abs/2112.04035
@@ArtemKirsanov Thank you, this looks very interesting, and I never knew this model existed.
I'll do a first read today, sounds very promising!
@@ArtemKirsanov A few days ago, I asked Claude+ "What does physically happen in your brain, when after having exerted a prolonged mental effort you feel unable to focus?" and look at points 4-7. They look very interesting in light of several of your videos.
Claude+:
Several things are likely happening in the brain when mental fatigue sets in after prolonged mental effort:
1. Depletion of mental resources: The brain requires glucose and oxygen to function properly. After prolonged mental activity, the brain's stores of glucose and oxygen can get depleted, leading to fatigue.
2. Buildup of waste products: As the brain is active, waste products such as lactate accumulate. Too much buildup of these waste products can impair brain function and lead to fatigue.
3. Neurotransmitter imbalance: Prolonged activity can lead to depletion or imbalance of neurotransmitters such as dopamine that are important for attention, focus and motivation. This can contribute to fatigue.
4. Decreased activation of attention networks: Parts of the brain involved in attention and focus, especially the dorsolateral prefrontal cortex, show reduced activation after prolonged mental activity. This makes it harder to sustain attention and concentrate.
5. Increased mind wandering: When mentally fatigued, there is more erratic activity across multiple brain networks, indicating increased mind wandering and distractibility. This makes it difficult to focus.
6. Altered brain wave patterns: There is a shift in the brain's electrical activity or brain waves, away from more alert patterns to slower waves as fatigue sets in. Slower, more irregular waves are associated with lower arousal and alertness.
7. Disrupted neural connectivity: Temporary mental fatigue may alter the levels of connectivity between different parts of the brain, disrupting optimal communication and coordination between brain networks. This impacts attention, focus and other functions.
With rest, all of these changes are reversible as the brain restores its mental and metabolic resources. Getting adequate sleep, taking breaks, and reducing cognitive load can all help prevent and relieve mental fatigue.
@@matveyshishov or you work through that mental fatigue and wait till your live does its damn glycolysis work.
Nah kidding this is great because I've been forcing myself to focus on this video right now because mental fatigue did set in.
But i also thing that it's attention span based, lien you say wandering off, i wander off seeking the next thing that would release more dopamine than previous.
So yeah it's depletion of Neurotransmitter but it's also based on learned imbalance, meaning other shit is more tempting and will release the dopamine needed to keep going isntead of focusing and having that delayed more longer reward pathway going.
This is why I focused on that video, and reading and writing this comment give me the closure i need for my brain to be more sure through a successful task that it was worth the forced focus amking more apt to have higher attention spans in the future. Do thanks for providing me that actionist with your comment. Haha
I believe that it would be more correct with regards to factorization that the categories are: (1) object qualities and (2) valuation/dynamics, not "sensory". All organisms use their senses to discern features or qualities, but then must relate these to valuation and dynamics within a context. The term sensory as used in this video would be imprecise within the discussion of factorization. How one "feels" about something is grounded on the recognized valuation within a context.
So the brain has learned to recognize the repetitive nature of spatial and non spatial variables in nature and in order to become as efficient as possible at processing information and solving problems critical for survival, it evolved the ability of factorization of different experiences and also evolved specialized cells to encode those variables in a "mental graph" and the ability to then combine different graphs into a higher dimensional graph that is specialized at solving the particular problem the brain is presented at any given circumstance. At least that's what I understand as a layman. Great video as always. I really like your content and although I'm not a neuroscientist myself I find the field more and more fascinating.
Exactly!! This is the essence.
I'm glad you enjoyed it ;)
crazy good video. best channel on youtube to explain neuroscientific topics. hope you will gain more attention
The hippocampal formation also includes the subiculum and the dentate gyrus, besides the hippocampus proper and the entorhinal cortex.
You, my friend, should be regarded as an international treasure. Im speechless at how great this video is in all facets i can think of right now. Thank you!
I feel a smart rat would smell the scent and have an ultra-violet trail to follow. Makes maze’s obsolete almost.
This video is amazingly clean and beautifully put together/choerent. I am appalled this video (and the rest of your videos) don't have millions of views (yet). Even if I have barely any hope of understanding most of what was said in this video, you came as close as possible to making me I think I do. And either way, if this entire video was in a completely foreign language with no subtitles, I would still have watched the whole thing for the visuals alone. Fantastic
Very nicely explained and interresting, just a little detail that I guess nobody noticed.. or was it on purpose to demonstrate what the brain is capable of.... at 3:45.. "neural represenations"...
I have subscribed and eager to learn more, thank you very much for this good work of sharing knowledge to everybody!!
Fascinating and visually stunning video, as always. Thank you Artem!
Keep up the work, you are one of the top science youtube channels!
Hi Artem, thank you for your fantastic videos! I really enjoy them and admire your ability to explain visualize such topic so elegantly and comprehensive! Brilliant visuals!
Small glitch at around 10:43 path integration: East arrow points West.
The only time TH-cam adds value to my life is when I'm watching Artem. The rest is attention and time theft.
Thank you 🙏🏾 for the premium quality content.
🎯 Key Takeaways for quick navigation:
00:00 🧠 El cerebro puede generar comportamientos flexibles generalizados en diferentes contextos, gracias a su capacidad de generalización.
02:13 🐀 En experimentos con ratas, se observa que tienen la capacidad de utilizar un mapa cognitivo interno para tomar decisiones en laberintos, incluso en situaciones nuevas.
04:06 🧭 La formación del hipocampo y la corteza entorrinal son cruciales para la construcción de mapas cognitivos en el cerebro.
06:07 🧭 Los tipos de neuronas, como las células de lugar y las células de cuadrícula, juegan un papel importante en la creación de mapas cognitivos.
11:13 🧲 Los mapas cognitivos se asemejan a estructuras de grafo que representan relaciones entre elementos del mundo.
14:53 🗺️ Los mapas cognitivos pueden expandirse para incluir información abstracta, como direcciones de giro en un laberinto.
17:08 🔗 La factorización de conocimiento en componentes estructurales y sensoriales permite al cerebro generalizar y simplificar la resolución de problemas.
19:57 🧩 Las células en el hipocampo y la corteza entorrinal ayudan a factorizar y conjugar información estructural y sensorial en los mapas cognitivos.
23:13 🗺️ Los mapas cognitivos organizan el conocimiento de manera eficiente, permitiendo la generalización y resolución de problemas en diferentes contextos.
Made with HARPA AI
this is quite possibly the coolest video ive ever seen on youtube
🎯 Key Takeaways for quick navigation:
00:00 🧠 Introduction to Cognitive Mapping
- The brain's ability to generate flexible behavior and generalize across contexts.
- Generalization explained through the concept of cognitive maps.
- Introduction to the two-part series on cognitive mapping.
01:39 🗺️ Cognitive Maps and Spatial Behavior
- Overview of the Tolman agenda machine and its role in solving the problem of cognitive mapping.
- Edward Tolman's experiment with rats in different mazes.
- Historical context of cognitive maps in spatial behavior.
03:53 🧭 Hippocampal Formation and Spatially Selective Neurons
- The role of the hippocampal formation in cognitive mapping.
- Introduction to place cells, grid cells, object vector cells, and landmark cells.
- Highlighting the diversity of cellular responses in the hippocampal formation.
06:07 🧩 Organization of Cognitive Maps
- Explanation of how entorhinal cortex and hippocampus work together.
- The general coordinate system provided by entorhinal cortex and the specific code from the hippocampus.
- Examples of spatial and non-spatial tasks influencing hippocampal neurons.
08:44 📊 Graph Theory and Cognitive Maps
- Introduction to graph theory as a mathematical formalism for organizing information.
- The importance of knowing one's location on the graph for effective cognitive maps.
- Path integration as a mechanism for updating one's position on the graph.
11:55 🌐 Unifying Spatial and Non-Spatial Domains
- Conceptual unification of spatial and non-spatial tasks using graph theory.
- The versatility of the brain in solving computational tasks across different modalities.
- The role of relational graphs in the brain's cognitive processes.
17:25 🧠 Latent Spaces and Factorized Representations
- Introduction to latent spaces and their importance in cognitive maps.
- The role of latent spaces in capturing abstract variables like sensory evidence.
- The concept of factorized representations for computational efficiency.
19:57 🤖 Generalization and Factorization in Cognitive Maps
- How factorization allows the brain to generalize knowledge.
- Comparing factorized and conjunctive representations in the hippocampal formation.
- Evidence for factorization in the responses of individual neurons.
22:05 🌐 Summary and Cognitive Mapping Applications
- Recap of behavioral and cellular evidence for cognitive maps.
- The role of cognitive maps in organizing knowledge efficiently.
- Teaser for the second part of the series focusing on building a computational system.
Made with HARPA AI
11:29 I know it's beside the point, but it's worth noting that human families, species taxonomy, and language taxonomy all differ from that idealized model in different ways when you try to understand them more accurately.
It's videos like this that support some personal theories about the world and natural laws, that touch on other theory's, and then leave me with a better respect for what it all is. Humbling myself in the pursuit of better understanding. Thank you.
I have to say the production of your videos is superb (I can't think of another word). Well structured and animated.
The only presentation tip I could give you, is to use short pauses on the conclusion and on important notes or exciting facts.
Otherwise perfect!
What are your thoughts on the Forward Model mechanism for planning and coordination? I'm wondering if you can also make a video explaining the relationship between cognitive map and FM. Thanks for making such amazing content!
Your videos are incredibly interesting and beautifully illustrated, makes 25 minutes feel like 5. Amazing stuff!
Thank you!
what was the point of this video, can you help me out seeming like he takes 25 minutes to say "its totally different for you when you walk around"
@@ryanbillie7708 I have no idea what you mean
Beautiful animation and explanation, you would make an amazing professor.
lovely visuals.
Hey, I appreciate you citing the sources in your description.
Artem, you absolutely set the bar for sci-comm on TH-cam. I truly haven't seen anyone better.
Thank you ❤
the title of this video really made me think you were gonna talk about the recent Tim Berhrens paper on HPC and mPFC representing the specifics and the generals of tasks… but nonetheless, nice video.
Great video, please keep it up! Would love to learn more about this and I especially appreciate you including sources!!
Thank you for a great video! Looking forward to the next part!
Thanks for presenting such a cool topic in such an accessible way! Your clear explanations and engaging visuals really made a complex subject easy to grasp. Fantastic work!
Thank you so much for making this video! It helped me finally understand what makes me so interested in neuroscience, AI - the principles that let things organize themselves. Great job, I’ll have to do some further reading!
Just fell in love with a new channel here. I love the graphics.
Tiny tiny typo at 21:40. Sorry for pointing it out. You're amazing!
When is the second part coming? Super interested. Any good reading material for modeling the latent space and factorization?
This is such an excellent summary of the field's findings, well done!. This is such an excellent summary of the field's findings, well done!.
Wow, what an amazing video, very well explained and simplified just in the right amount so that we can understand it but we know there's a lot more about the subject to be studied.
Great job man, these videos breaking down how we learn at a fundamental level are really helpful for creating better studying techniques.
Thank you and keep it up!
I'm very late to the party I see, but I'd just like to congratulate you on an excellent educational AND entertaining video; I'm highly impressed! Not only do you 1. use simple but beautiful and engaging animations, 2. you have a "story to tell" from A to Z (well ok there's a part 2 but I'm going there next), 3. you use great analogies, 4. explain jargon/terminology, you 5. summarize in between and at the end (spaced repetition!) and 6. your voice adds to the visuals instead of simply copying. 6 out of 5 stars from me, and an immediate subscribe! Best of luck to you, you really have a talent for education! (See, i factorized my critique, proving i actually learned something ;-)
Excellent production. I thoroughly enjoyed this. Everything about the brain is so fascinating.
This is such an excellent summary of the field's findings, well done!
This video was fantastic! The visual models really helped me get a good grasp on all of these unfamiliar concepts. Thank you!
Coming from a computer science background I'm relating all this to a CPU. The hippocampus is like an ALU, dedicated hardware to accelerate certain tasks. Is there a structure in the brain analogous to the control logic? Something that directs all this "hardware" to perform the task
It's crazy how similarly computers and brains work🙆🏾♂️
absolutely stunning work. Congrats once again! I think you have found your calling.
Two things that came to my mind.
1. This video reminds me of when I studied Kant's Aesthetics and philosophy in general during high school.
2. Now I wonder what goes on in a mathematician's mind when he's writing a proof to a theorem, or a surgeon applying his knowledge to a case he's going to operate.
Nice! I was asking about just this in one of your previous videos!
Wow, very interesting video! The visuals help follow along. Great job!
Hexagonal grid: Hmm, is it really hexagonal, or simply triangular, as depicted in the graphics. Obviously a triangular grid contains hexagons, but the difference is that a hexagonal grid is missing some nodes (the centers of the hexagons) compared to a triangular grid.
I had a weird experience a couple of years ago where I was out for a walk and I suddenly felt like I was on a street which I had completely forgotten about for 7 years, in a different country. Could that have been caused by the relative placement of houses and roads suddenly triggering some place cells and grid cells in such a way that brought that all back?
Great video, It has increased my curiosity in computational neuroscience...
Amazing visualizations. As someone looking to teach through youtube, I'm definitely taking notes.
This ranks as one of your top three videos. Bravo.
Wow, thank you!
This channel is what i have been looking for fr
your logo is oroboros/infinity inside the frame of a human brain,yes?
fantastic video! so much new details i gad no idea about. thanks for sharing,man!
Very cool video. On 10:44 the arrow of 'East' is drawn equivalent to the 'West' arrow
I was wondering if I was the only one to notice that or if I was hallucinating.
10:41 East and west is in same direction.
Btw amazingggggg video, keep it up.
I can't wait to go through all your videos. Nice intro to these subjects. The one thing that I keep admiring is the visualisations. Are you producing these things alone or with help? How long does it take to create and put together all the visualisations for the video?
i am wondering as well
Exceedingly professional work!
Wow. It's an amazing work. Thank you for sharing it with us!
One of the few videos on TH-cam where the description contains links to scientific publications of different authors.👍