Talk:Hebbian theory

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Neuroscience (Rated Start-class, Top-importance)
WikiProject icon This article is within the scope of WikiProject Neuroscience, a collaborative effort to improve the coverage of Neuroscience on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
 Top  This article has been rated as Top-importance on the project's importance scale.
 

Merging with Hebbian learning[edit]

Hebbian learning is encompassed within Hebbian theory. Hebbian learning should be incorporated into this article. Thoughts? Semiconscious (talk · home) 06:58, 25 December 2005 (UTC)

people looking Hebb's rule up on the net, will most likely search for Hebbian Learning - as it is refered to in textbooks, too. I have so far not run accross the term Hebbian Theory, but Hebbian Learning very regularly during my research in neuroscience. Thus, I suggest to at least make a forward from the page "Hebbian Learning" to "Hebbian theory" or even forward the other way round - which though probably doesnt make much sense since the latter is more encompassing. Merging makes at the current state much sense since this page is very slim. --AnotherWorld 20:52, 27 February 2006 (UTC)

Agree; merging seems sensibleGleng 21:26, 4 March 2006 (UTC)

Yes, merge them. wmaier, 11 March 2006

Yes to merge. adamg, 21 April 2005

Yes, merge, but please make sure to have those forwards suggested by AnotherWorld Limbonic 09:33, 23 April 2006 (UTC)

Yes to merging Hebbian Learning into Hebbian Theory. I'd try to do it myself, but I think I'd screw it up. Anyone else want to go for it? Doc Pato 09:42, 1 June 2006 (UTC)

No to merging. Hebbian learning as an article on a type of artificial neural network is rather small and could do with more graphics, but as an idea for software, it is distinct from the theory about how the biological brains work. One article is about biology and cognition, the other about computers. -RN 20. july 2006

Life is short. merge the articles. include sections on how hebbian theory has been applied to neural networks, software and biological systems. how many Hebb's do we need? October 3 '06

Hebbian Learning in itself is not complete. It would be a good idea to merge it with Hebbian Theory. Additionally, the search term "Hebbian Learning" should be able to point the user to "Hebbian Theory". If this can be done, then I am for the merging. -- Gautam

No to merging. Hebbian learning has evolved into a mathematical concept dealing with networks and software, not how Hebb envisioned it. His theory is biological and more specifically neuronal in function. Learning is a sharp adjunct. They don't belong together. To AnotherWorld, in researching for a neuroscience review, I've come across Hebbian Theory exclusively, with no mention of Hebbian Learning. Both entries can see an increase in content, however. This would make the merger completely unnecessary as well. lucash, 28 Oct 2006.

Different ideas[edit]

Neuronal assemblies and Hebbian learning are very different ideas created by Donald Hebb, which are related sometimes but each is a different concept. Because both associated with name of "Hebb" it doesn't mean they are the same thing. Hebb's writings are in biological context. All those biological discussions deserve to have another seperate article. Sohale (talk) 13:59, 3 August 2008 (UTC)

Flagging inappropriate links[edit]

Both Hebbian Learning and Hebbian Theory have links to a purported tutorial. This tutorial only works if you purchase an expensive piece of proprietary software. This seems inappropriate.

--Fred —Preceding unsigned comment added by 137.43.155.87 (talk) 10:13, 11 November 2009 (UTC)

Modifications and meaning of Hebb[edit]

The last sentence of the article is not really right. While simple Hebbian learning is unstable, the other rules that are cited (BCM, Oja etc) are still just modifications of the basic Hebb rule. The whole point of Hebbian learning is that it allows certain connection strengths to grow exponentially, somewhat in the way that organismal reproduction can lead to exponential growth. In an real situation, exponential growth will run out of steam and if this slowing of growth affects different connections equally, it will not change the underlying selective growth, although it will stabilize overall growth. If one had a pure Hebb rule in the brain it would work quite nicely, because something will always prevent unlimited growth. The BCM rule also introduces a nonlinearity into the rule (which Oja does not), but this is really quite separate from the competitive aspect. The article does not really explain why Hebbian learning is so important. The key idea is that it makes the overall growth of a connection sensitive to the various correlations between the inputs. This is because the growth depends immediately onto the correlation between the activity of the input to a connection and the output activity, which in turn reflects the combined effects of all the inputs, not just the input to the weight being changed.Paulhummerman (talk) 15:58, 27 November 2010 (UTC)

Inverse Hebbian theory[edit]

I am researching Post Operative Cognitive Disorder (POCD), ICU psychosis and its possible relationship to Hebb's postulate or more accurately the inverse of Hebb's postulate. It is now taken for granted that "neurons that fire together wire together" Numerous scholarly articles show that the opposite is also true. This being that when Neurons/Engrams are disconnected they tend to stay disconnected. Benzodiazepines such as midazolam (gaba modulator) seem to be a common denominator in POCD, ICU psychosis, and what is referred to as paradoxical reactions in procedural sedation, formerly referred to as conscious sedation. These problems have existed since the dawn of modern anesthesiology and are so closely linked that they are clouded by the problem of familiarity. This condition makes it extremely difficult to identify cause and effect. I have a theory that may help identify those most at risk.

I call it The Missing Domino Effect.

Instead of a CSI representation of synaptic activity, simplify it to a fancy set of dominoes. The dominoes falling represent the inter-neuronal connections. The first domino hits the second and it hits three more and branches into a cascade. Imagine now that the second domino is removed. Game over. This describes the action of midazolam, its analogues and even alcohol. They all increase GABA binding and inhibit synaptic activity globally, but the first neurons to fall are the neuronal connections in the prefrontal cortex.

Therefore, logic dictates that the more activity there is in the prefrontal cortex, the less dosage it will take to shut it down. Recent research also indicates that alcohol affects GABA modulation in a very similar way to benzodiazepines. I suspect that people having trouble handling one glass of wine will also need decreased dosage of midazolam.

This points to the extreme variability of mental function from one subject to another. In the field of psychology, it is referred to as executive function, and it's constantly changing.

In this case, the bigger (cognitively speaking) they are, the easier they fall. It would take functional imaging of the prefrontal cortex, and a few simple experiments to determine this.

Does anyone have any ideas on the subject?

Vernon su@vpw.us

Vwilkinsjr (talk) 16:06, 23 August 2012 (UTC)