Discussing the article, "Understanding Memory in Deep Learning Systems: The Neuroscience, Psychology and Technology Perspectives" by Jesus Rodriquez

Our memory is a huge important part to our daily lives. From remembering the notes you learned in your biology class to remembering how to walk, our memory controls what, how, and why we are doing something. Some memories don’t last for very long which is why our memory isn’t like a video camera that captures every detail of the event; for example, the color of your friend’s shirt. Certain memories can go from our short-term memory to our long-term memory storage. However, there are different explanations (theories) to how we process information and store it away to remember later. The two theories are based around two kinds of psychology, neuroscience and cognitive psychology.

According to the neuroscience theory of memory, there are three parts that help move the information we encode to our long-term storage. The first part is the thalamus which encodes the senses we perceive (what we see, hear, touch, taste, smell) and then moves that information to the appropriate lobe. Next, the processed information arrives at the prefrontal cortex where it turns it into short-term memories that only last for a couple seconds unless taken to the last part, the hippocampus, to store into long-term memory. However, the memory is broken down into different pieces that go to different cortices of the brain. This uncertainty of how the different pieces come back together to create a cohesive memory is called “the binding problem”.

The binding problem has been explained with a couple different possibilities. One theory being that the pieces of the memory are connected to electromagnetic vibrations. These connections help create the memory back to one so we can remember it as a whole. The other theory of the binding problem explains more how those memories get recalled which is based on the cognitive psychology theory of memory, called the priming effect. The priming effect is when you think of a word or hear something and it triggers another word or memory; for example, when you think of the word ‘school’ it may trigger a memory like sitting in class or a more specific memory related to your experience with school.

Memory is a very important and interesting concept because we use it all the time. I’m using memory (unconscious memory) right now to remember how to spell and type. Many stimuli can trigger memories of events that happened years ago. For example, the fall weather can take you back to when you were younger and jumping in the big pile of leaves you just raked. I think our memory is very helpful but at the same time it isn’t always that reliable. False memories can occur when trying to remember specific details, but certain words, due to the priming effect, may in fact recall those details and store them back in long-term memory to remember for an unlimited amount of time.

References

Rodriguez, J. (2018, July 25). Understanding Memory in Deep Learning Systems: The Neuroscience, Psychology and Technology? Retrieved from https://towardsdatascience.com/understanding-memory-in-deep-learning-systems-the-neuroscience-psychology-and-technology-78922ae6a1dd