Scientists from the Salk Institute have made a breakthrough in understanding how the human brain works. Researchers have proven that neural waves propagating across the visual cortex are not just background noise, but a complex computational mechanism. It is precisely these electrical impulses that allow the brain to construct internal representations of the three-dimensional external world.

Previously, it was believed that the visual system operated according to a strict hierarchy of signal transmission. However, new research shows that the brain uses "traveling waves" to perform the spatiotemporal computations necessary for predicting events and accurately perceiving reality.

From Searching for Glasses to Generative Models

The problem of perception was described back in 2020, when a team led by neurobiologist John Reynolds noticed a strange correlation. Scientists found that traveling waves in the visual system directly influence the ability to notice objects located right in front of the eyes. This explained the phenomenon where a person searches for an item lying in plain sight for a long time: if the wave pattern in the brain does not match the object's position, it remains unnoticed.

In the new study, Reynolds and his colleagues expanded on this theory. They hypothesized that the neural connections generating these waves constantly change their structure, adapting to the external world. Scientists draw a direct parallel to the operation of large language models (LLMs) in artificial intelligence.

Just as neural networks learn the statistical structure of data to generate a meaningful response, the brain uses traveling waves as a biological generative model. It is created "from scratch" based on sensory experience, allowing the brain not just to record an image, but to actively interpret it.

Four Functions of Neural Waves

The study described in detail the role of these dynamic patterns in the visual cortex. The waves perform four critically important functions that allow us to interact with the world:

  • Modulate perception from moment to moment, changing sensitivity to signals.
  • Transform fresh visual data into coherent images.
  • Generate short-term predictions about the environment, allowing for the anticipation of events.
  • Store and reproduce memory patterns over time.

Unlike classical systems focused on simple information transmission, these processes retain the "history" of stimulation. This allows the brain to build predictions based on the connection between the time and place of previous events.

A Global Brain Mechanism

Dynamics similar to neural traveling waves (nTW) are a fundamental characteristic of brain function in many different species—from fish to humans. These processes manifest at various spatial and temporal scales.

In different parts of the brain, waves perform specific tasks:

  • In the motor cortex, waves in the beta range serve as a predictor for the onset of movement.
  • In the prefrontal and parietal cortex, similar processes are linked to mechanisms maintaining working memory.
  • Beyond the neocortex, these waves play a key role in memory encoding and organizing signals for reinforcement learning.

At the level of the whole brain, this dynamics is most pronounced during sleep, highlighting the global nature of wave processes in ensuring cognitive functions.

A New Dimension of Neurobiology

Experimental data confirms that neural traveling waves add a "third dimension"—spatiotemporal computations within individual areas—to traditional hierarchical models of brain function. Impulses can propagate continuously or discretely via long horizontal fibers connecting neurons.

These processes directly control perception and behavior, modulating neural excitability in real-time. The functional convergence of biological processes with the operation of artificial intelligence models opens new horizons for understanding how a living organism creates a flexible internal representation of the external world.