What is Hebbian Learning?

What happens in our brains when we learn something new? What changes do neurons and synapses go through in the formation of new memories? In 1949, Donald Hebb proposed that learning and memory occur through a process that can be summarized as “neurons that fire together connect together”. The expression “together” is not used to mean that two neurons, A and B, are excited exactly simultaneously. The point is that A does not fire at the same time as B, but a little before each time, as if there is a causal relationship. We can liken this to the formation of a path that becomes a little more obvious with each passing of the road. According to the Hebbian learning theory, each time two neurons communicate, the connection between them becomes stronger. Hebb’s work also pioneered the concept of action potential (spike) timing dependent plasticity (STDP).

Hebb’s theory was first applied to experiments on conditioning designed by the Russian scientist Ivan Pavlov. Conditioning is often a learning process. For example, when a bee is given sugar, a certain neuron fires and the bee’s tongue sticks out. If we expose it to the smell of lemon before giving sugar and repeat this several times, the bee can stick out its tongue just because of the smell of lemon, even if the sugar is no longer given. The neuron is thus fired only by smell. Hebbian learning strengthened the connections between neurons. The power of Hebb’s idea comes from the demonstration of changes that occur at the molecular level in interneuron connections. When we learn a new word, such as “Hebbian,” we create new connections in language-related brain networks to the Hebbian learning pathway.

Molecular Basis of Learning

The gap between two neurons that connect with each other is called the synaptic cleft. By means of neurotransmitters that use this gap as a bridge, the first neuron can excite or suppress the second. Learning occurs as a result of the combination of these two types of interaction. Evidence for the mechanism of learning has been demonstrated even in simple creatures such as the sea snail. For example, the glutamate neurotransmitter released from the first neuron crosses the synaptic cleft and binds to its receptors in the second neuron. This binding creates an increasing effect on glutamate receptors in the second neuron. The increase in glutamate receptors enables the second neuron to be stimulated more easily in subsequent signals.

Hebbian learning or associative learning states that the simultaneous activity of cells strengthens the synaptic connections between these cells. It also offers a biological basis for memory rehabilitation with error-free learning methods. In neural network studies of cognitive function, it is generally accepted as the neuronal basis of unsupervised learning.

Engramlar

Engram is a theoretical concept defined as a unit of information stored in the brain by biophysical or biochemical changes in response to external stimuli. Clarifying the actual formation mechanism and place of engrams is an important subject of neuroscience. Some researchers have suggested that there is no specific memory-related region in the brain and that information is scattered throughout the cerebral cortex. However, the prevailing view today is that memory for complex tasks is distributed across different neural systems, but specific areas exist in the brain for certain types of information. Sometimes even a single neuron may be responsible for storing certain information. Learning and memory are still not very well understood. However, the cerebral cortexWe know that structures such as the striatum, hippocampus, amygdala, and cerebellum play an important role in memory.

Read More  What is the Limbic System? Duties

Hebbian theory deals with how neurons connect to each other to turn into engrams. If one cell constantly stimulates the other, synaptic nodes develop from the first cell to the second or existing ones grow. Eric Kandel’s work provided evidence for the existence of Hebbian learning mechanisms in the synapses of marine invertebrates. It is much more difficult to do similar studies on the nervous systems of vertebrates because the number of variables that need to be controlled is large.

Effect on Artificial Neuron and Neuron Network Models

From the perspective of artificial neurons and neuron networks, the Hebb principle can be used to determine how to change the effects (weights of each other) of model neurons. If two neurons are activated simultaneously, the effect between them increases, and if they are activated separately, it decreases. Nodes that are both positive and negative at the same time have a strong positive effect, while opposing nodes have a strong negative effect. In summary, the Hebb principle is one of the basic principles used in the development of artificial intelligence. However, Hebb’s rule is not stable in any neuron model. For this reason, neural network models often use other learning theories such as BCM theory, Oja’s rule, or General Hebbian Algorithm.

While Hebbian models are often used in long-term potentiation, there are some exceptions and the theory may be too simplistic in some ways. One of the most well-known exceptions is that synaptic modification does not occur only between A and B neurons, but other neighboring neurons are also involved. The Hebbian modification is dependent on retrograde signaling for the pre-synaptic (presynaptic) neuron to change. Nitric oxide is the most common neurotransmitter that carries out retrograde signaling. This highly soluble gas can also affect other neurons in the vicinity. The concept of diffuse synaptic modification can complement the traditional Hebbian model.

Read More  Brain Development and Influencing Factors

Hebbian Learning and Mirror Neurons

Hebbian learning has also been used to explain mirror neurons. Mirror neurons are neurons that are activated both when a person performs a certain action and when they see or hear someone else do a similar action. Mirror neurons are effective in understanding other people’s actions and empathizing. Visual, auditory and tactile stimuli received while performing a certain movement associate the movement with the senses. As this experience is experienced, the connection between them becomes stronger and only by seeing or hearing the movement, motor (movement) neurons can be stimulated and mirror neurons can be formed.

Related Posts

Leave a Reply

Your email address will not be published.