Strange Dreams Might Begin the Brain's Learning

Strange Dreams Might Begin the Brain's Learning ...

According to a new study by researchers from the University of Bern, Switzerland, dreams - especially those that simultaneously appear realistic, upon a closer look, bizarre - help our brain learn and extract generic notions from previous experiences. The study, which was developed within the Human Brain Project and published in eLife, demonstrates how dreams are perceived, using machine learning techniques and brain simulation.

The need for sleep and dreams to develop and develop skills has long been emphasized, and the impact that a single restless night can have on our cognition is well known. What we lack is a theory that combines this with a combination of experiences, a generalization of concepts, and creativity, according to Nicolas Deperrois, the lead author of the study.

During sleep, we experience two kinds of sleep cycles, one after the other: non-REM sleep, when the brain replays the sensory stimulus experienced while awake, and REM sleep, where spontaneous bursts of intense brain activity produce vivid dreams.

In order to introduce an element of unusualness in artificial dreams, researchers took inspiration from a machine learning technique called GANs. In GANs, two neural networks compete with one other to obtain new data from a similar dataset, in this case a series of simple images of objects and animals. This procedure creates new artificial images that can look superficially realistic to a human observer.

During wakefulness, the researchers simulated the cortex in three sections: wakefulness, non-REM sleep, and REM sleep. During wakefulness, the model is exposed to photographs of boats, cars, dogs, and other objects. In non-REM sleep, the model is exposed to some occlusions. A simple classifier is used to show how easily the identity of the object (boat, dog, etc.) can be read from the cortical representations.

As our model learns, non-REM and REM dreams tend to creatively combine these experiences. Interestingly, when the REM sleep phase was suppressed in the model, or when these dreams became less creative, the accuracy of the classifier decreased. Similarly, when the NREM sleep phase was removed, these representations tended to be more sensitive to sensory disturbances (here, and occlusions).

According to Deperrois, wakefulness, non-REM and REM sleep are all associated with learning, such as using the stimulus, instilling that experience, and clarifying semantic concepts. This phenomenon should not be surprising as dreams become bizarre. The next time youe having crazy dreams, perhaps don''t try to find a deeper meaning - your brain may be merely organizing your experiences.

You may also like: