The use of these labels to isolate and protect information in the brain has already been shown. For example, when monkeys are preparing to move, the neural activity in their motor cortex represents movement but they do so in such a way as to avoid distractions and the signs governing specific rules in the muscles.
However, it is not clear how neural events are altered in this way. Buschman and Libby want to answer this question from what they see in their mouse barn. “Once I started in the lab, it was hard to imagine how such a thing could happen with neural shooting events,” Libby said. He wanted to “open the black box of what neural networks are doing to create this disruption.”
In experiments they could, they saw the possibility that the blood components of the ears that contribute to the representation and memory. Instead, it was shown that the same number of neurons is affected, and that the activity of neurons can be effectively divided into two groups. Some were “steady” in their behavior during visualization and memory, while other “switches” reversed their responses to each use.
Surprisingly, the combination of stable and modified neurons was enough to process the information and convert it into memory. “That’s all magic,” Buschman said.
Instead, he and Libby used mathematical methods to show that the machine was the best way to create a dynamic shape and memory: It requires fewer neurons and less power than other methods.
The results of Buschman and Libby stem from a scientific theory: that neurons, even in small parts, produce more powerful signals than previously thought. “These parts of the cottage that are low on food also have some interesting ingredients that we may not have appreciated yet,” he said. Miguel Maravall, a psychologist at the University of Sussex who did not participate in the study.
This work can help to reconcile the two sides of the ongoing debate if short-term memory is maintained through constant, steady or through solid neural numbers that change over time. Instead of just coming from one side or the other, “our results show that they were all right,” said Buschman, with stable neurons that complement the original and change neurons. Its integration is useful because it “helps to prevent distractions and create this circular exchange.”
Buschman and Libby’s research may be appropriate in more representative areas. They and other researchers are hoping to look at circulating machines in other ways: how the brain stores multiple thoughts or goals at once; how it works in the process of disruption; as representing internal components; how it controls perception, combined interest.
“I’m very happy,” Buschman said. Looking at the work of other researchers, “I remember seeing, there is a stable neuron, there is a change in the neuron! You see them everywhere here.”
Libby is interested in the meaning of their results in spyware, especially in the design of useful AI networks that need to do more. “I would like to see if the people who set up the neurons in their neural networks to be stable and switch properties, instead of useless things, have helped their networks in some way,” he said.
All in all, “the results of such letters will be very important and interesting to know,” Maravall said.
First article reprinted with permission from Quanta Magazine, an independent document of Simons Foundation whose job it is to promote the understanding of science by explaining what has happened and to investigate mathematics as well as physical and life sciences.
Many Great Stories