Tech

Neural noise reveals uncertainty in our memories


In this moment Between reading a phone number and tapping it into the phone, you may find that the digits have mysteriously gone wrong — even when you’ve written the first digits into your memory, the the last digit can still be blurred uncountable. Number 6 before 8 or after it? Are you sure?

Retaining such trifles long enough to act on them is based on an ability known as visual working memory. For years, scientists have debated whether working memory only has space for a few items at a time or if it has limited room for detail: Perhaps our minds’ ability to experience spread over a few clear memories or a multitude of more questionable fragments. .

According to a recent newspaper in Neuron from neuroscience researchers at New York University. Using machine learning to analyze brain scans of people engaged in a memory task, they found that the encoded signals estimated what people thought they had seen — and the statistical distribution of noise in the signal encodes the uncertainty of the memory. The uncertainty in your perception may be part of what your brain shows in its recollections. And this sense of uncertainty can help the brain make better decisions about how to use its memories.

The findings suggest that “the brain is using that noise,” Clayton Curtisa professor of psychology and neuroscience at NYU and the author of the new paper.

The work adds to growing evidence that, even when people don’t seem adept at understanding statistics in their daily lives, the brain often interprets the sensory impressions of the brain. it about the world, both present and remembered, in terms of probability. Insight provides a new way to understand the value we assign to our perception of an uncertain world.

Predictions based on the past

The neurons in the visual system work in response to specific sights, like a perpendicular line, a particular pattern, or even cars or faces, causing sparks to the part. rest of the nervous system. But on their own, individual neurons are noisy sources of information, so “it is unlikely that individual neurons are the currency the brain is using to infer what it sees.” see,” said Curtis.

For Clayton Curtis, a professor of psychology and neuroscience at New York University, recent analyzes show that the brain uses noise in its neuroelectrical signals to express uncertainty. on cognition and encoded memory.Courtesy of Clayton Curtis

Most likely, the brain is combining information from a population of neurons. So it’s important to understand how it works. For example, it could be averaging information from cells: If some neurons fire most strongly when seen at 45 degrees and others at 90 degrees, the brain can balance and average their input to represent a 60-degree angle in the eye’s field of view. Or perhaps the brain has a winner-takes-all approach, with the most active neurons being taken as indicators of what is perceived.

“But there is a new way of thinking about it, influenced by Bayesian theory,” says Curtis.

Bayesian theory – named after its developer, the 18th-century mathematician Thomas Bayes, but later independently discovered and popularized by Pierre-Simon Laplace – incorporates uncertainty into the way its probability approach. Bayesian inference deals with the degree of confidence one can expect an outcome to occur given known circumstances. When applied to vision, that approach could mean that the brain understands neural signals by constructing a function of likelihood: Based on data from previous experiences, these Which point is most likely to have created a certain trigger pattern?

Wei Ji Ma, a professor of neuroscience and psychology at NYU, has provided some of the first concrete evidence that populations of neurons can perform optimal Bayesian inference calculations.Courtesy of Wei Ji Ma

Laplace realized that conditional probability was the most accurate way to talk about any observations, and in 1867 the physician and physicist Hermann von Helmholtz connected them with the calculations that their brains could make. we can do in the cognitive process. However, very few neuroscientists paid much attention to these ideas until the 1990s and early 2000s, when researchers began to discover that people were doing something like thinking. probabilistic reasoning in behavioral experiments, and the Bayesian method is beginning to prove useful in a number of cognitive and motor control models.

“Everybody who started talking about the brain was Bayesian,” says Wei Ji Maa professor of neuroscience and psychology at NYU and a new professor Neuron author of the paper.



Source link

news7g

News7g: Update the world's latest breaking news online of the day, breaking news, politics, society today, international mainstream news .Updated news 24/7: Entertainment, Sports...at the World everyday world. Hot news, images, video clips that are updated quickly and reliably

Related Articles

Back to top button