Published on: January 3, 2025
Know s’adapter and take decisions facing unforeseen is essential in our everyday lives, but how does this translate into our brain ? French researchers publish their results in the prestigious journal Sciences Advance allowing us to provide some answers to this question.
Even when we are faced with the same situation several times, our brain never always activates in the same way: brain activity shows variability. These small variations which can be considered as “noise” would play an important role according to this study: they would help the brain has s’adapterto explore different solutions and to react better to complex or uncertain situations.
To test this hypothesis, the researchers used innovative tools used in artificial intelligence : artificial neural networks inspired by brain human. They were thus able to program two models: neural networks always reacting the same way without variability in their activitywhile others were modified to simulate variable brain activity. When confronted with new and uncertain situations, networks with variability are flexibly adaptedwhile the rigid networks showed a inappropriate behavior.
These results suggest that the Brain variability plays a key role in the ability to adapt and adjust decisions based on new information or changes in the environment, particularly in unforeseen contexts.
This study raises the question of the impact of cerebral variability on symptoms which disturb decision making, especially in situation of uncertaintyas in the schizophrenia or the obsessive compulsive disorder. By measuring this variability in patients, scientists seek to better understand the underlying mechanisms.
-Sources :
- “AI: better understanding decision-making thanks to artificial neurons”, Presse Inserm
- Charles Findling, Valentin Wyart, Computation noise promotes zero-shot adaptation to uncertainty during decision-making in artificial neural networks.Sci. Adv.10,eadl3931(2024).DOI:1126/sciadv.adl3931
Neuroscientists use AI-based computational models to simulate brain function and understand its cognitive processes. These models, inspired by biological neural networks, make it possible to test hypotheses on the processing of information by the brain and to study complex functions such as perception, memory, learning and decision-making.
Decision making is an essential cognitive function allowing us to choose an option among other alternatives based on different factors. This function can be impaired in several brain diseases, particularly in psychiatric disorders. It is thanks to research that the understanding of the brain and its dysfunctions advances!
> Learn more about donating
Related News :