Is there a unit of information in neurology yet...

Is there a unit of information in neurology yet. I want to know if specific memories can be linked to specific neurons or neural networks.

hasn't been done yet

We still don't know if memories are stored in one neuron, or a small number of neurons, or even distributed over a large number of neurons.

nigga what, memories are highly abstract and dispersed (in terms of neural structure) what are you on about. You don't speak in "neurons" for such things.

It's neuroscience/neurobiology not neurology and what? You actually think non-biological baseless abstractions like "information" and "memory" have 1 to 1 or even any useful correspondence to reality?

>nigga what, memories are highly abstract and dispersed
that's a physicist point of view, because that kind of system where everything relates to everything else is the kind we know how to solve best, but no, what you say isn't certain at all

Most people think neurons encode coincidences and features in events (spike timing dependent plasticity). At any given time, something on the order of 1% of higher cortical neurons are firing, which is a vast number. If the brain operates in any way similarly to CNNs (which are visual cortex inspired), then in most cases a neuron will either fire when a specific feature is present in sensory inputs, or when it is activated by a neighbor due to previous associative learning. So for instance, the memory of your mother's face might be encoded by millions of neurons, each of which is sensitive to some small feature like nose width, hair color, etc. Also those units are probably triggered by people with similar features, its just that you have some pattern completion systems like the hippocampus that bind it all together when the concept of your mother is being recalled either by seeing her or something reminding you of her.

So how do quadriplegics know stuff if the sensory information can't get to the hippocampus?

Why not?

What do you mean? Sensory information includes information from every single nerve cell embedded in your body e.g. eyes, nose, mouth, ears, skin, etc. In fact it seems the brain is pretty much 'driven' by sensory information. People where the prefrontal cortex is too dominant in relation to sensory areas have massive issues (autism), problems with synchronization seem to be implicated in schizophrenia, and sensory deprivation often leads to mild insanity.

Unrelated tangent, but if you look at charts of what different areas of the brain supposedly do, areas between specific sensory areas usually encode information which utilizes both areas. E.g. the pre-motor cortex integrates visual information to guide movements, and Broca's area (responsible for speech production) is between the frontal cortex (complex planning) the motor cortex, and the primary auditory cortex.

Ah ok I got you. I'm a brainlet when it comes to neuroscience and I'm learning so far (independent learning) that the somatosensory cortex is important for sensory input but that there are lots of cortex eg: Visual Cortex by the occipital lobe which is used for actual seeing (in contrast to sensations on the physical eyeballs, I think? when it comes to sensations of the eyes in the somatosensory cortex).

This shit:

Yes exactly, the somatosensory cortex sort of represents a miniature version of your whole body related to tactile sensation coming from different nerves in your body. This also maps on to another 'miniature map' of your body in the motor cortex which corresponds to voluntary movements of different limbs, often driven by inputs of your reward and planning systems (amygdala, striatum, prefrontal cortex, etc) in conjunction to visual and other feedback from other areas.

Yeah it's very interesting this stuff, I'm just learning the conscious tracts at the moment. Crazy how complicated this stuff is.

Also produces this retard.

I get the feeling that you did a degree in this right?

What is your occupation, do you do research? It is evident that you know what you are talking about, would like to read more.

Nah I taught myself, my degrees were in comp sci + artificial intelligence and the computational modelling of associative learning.

Just finished a PhD, looking for post-docs. I recommend reading introductory text-books on neuroscience, the basics of associative learning (e.g. scholarpedia) and reinforcement learning (Sutton's Reinforcement Learning is available online), how modern artificial neural networks work broadly (CNNs, GAN, recurrent nets, predictive coding,backpropagation), look into cortical models of learning, spike timing dependent plasticity, how neurotransmitters work (e.g. correlation of dopamine with prediction errors i.e. Schultz etc.), interesting clinical phenomena which reveal how the brain functions (change blindness, schizophrenia, blindsight, phantom limbs, parkinsons), the effects of hormones on the brain (testosterone, oxytocin, vasopressin), the connection between cortisol, inflammatory cytokines and depression. nootropics related to acetylcholine, dopamine, BDNF and their influence on neuronal firing and axon growth, the relation of neuronal pruning in artificial and biological neural nets and how they relate to overfitting/generalization. There's a lot to take in, so its better to focus on having a broad overview rather than trying to memorize something very specific like brain areas or the receptor binding potential of molecules.

Ah ok.

Random question, why doesn't the dick have tons of sensory receptors? Why is it small compared to the hands? Doesn't it make sense to have lots of innervation there?

Probably because you need a lot of nerves there to incentivize sex and disincentivize damaging your dick, but having more neurons process that data isn't as important. Your hands are over-represented like that because its integral to using your hand for fine motor skills. Your dick doesn't really need fine motor skills to the same extent. Also you use your hands all day long, while you usually will have sex perhaps once a day for 10 minutes. Just a guess.

Thanks!

What can you say about Hawkins's hierarchical temporal memory and memory-prediction framework concepts? Do they seem plausible to you? Seems that they successfully integrate several observations about brain function.

Also found the Tensor network theory quite interesting, although explaining other than motor functions with that is going to prove difficult.

Sorry for rambling. MD/PhD 5th year, researching DTI use in DBS planning, aiming for the broader view that you mentioned.

Yes?

isn't there even a solid hypothesis that there is one new spine per new memory?

Yeah I definitely think that at least in terms of cortical function, predictive coding and free-energy based principles are going to be important, like Hawkins' work. Though I personally believe that associative learning (esp. w.r.t. to the hippocampus) and aspects of a 'localized' (i.e. passing the teaching signal along incrementally down a hierarchy of neurons) backpropagation will feed into it as well. Not familiar with the tensor network theory, I'll look into that.