Summary: A recent research reveals that both people and macaques continue to modify their behavior yet after developing a working strategy. They gently alter their responses based on their most recent encounter, rather than sticking to a well-known powerful approach. This continuing investigation may help them adapt to changing conditions or learn more effective techniques over time.
Researchers discovered that subjects used a threat design to predict event timing but also adjusted based on previous studies. The results suggest that inquiry is a natural course of action in decision-making, even when it leads to pointless problems. The study also confirms the similarity of marmosets to mortal cognitive processes, supporting the use of them as autism models.
Important Information:
- Exploration Tendency: Yet after mastering a task, humans and marmosets change how they behave based on previous failures.
- Adaptive Learning: This behavior may enable animals be flexible in uncertain conditions.
- Results for Autism Research: Marmosets are used as models to study forecast differences in dementia.
Origin: Picower Institute at MIT
Sometimes it’s a lifestyle steal or a responsibility, or a little of both.
A surprising finding in a recent MIT study may point out that people and animals equally have an inborn propensity to constantly update their approach to a task also when they have already learned how to do it and even if the deviations occasionally lead to unnecessary errors.
The behaviour of “exploring” when one could really be “exploiting”, may produce feeling for at least two factors, said , Mriganka Sur, senior author of the study published Feb. 18 in , Current Biology.
In this ambiguous world, changing behavior from the ideal state every so often had support show needed adjustments because just because a task’s rules seem set one moment doesn’t mean they’ll stay that way.
Additionally, trying new things when you already have an understanding of what you enjoy will reveal potential alternatives to the good thing you’re doing at the moment.
Sur, Paul and Lilah Newton, Professor at The Picower Institute for Learning and Memory and the Department of Brain and Cognitive Sciences at MIT, said,” If the goal is to increase reward, you should never stray once you have found the perfect solution, but you keep exploring.”
” Why? It’s like foods. We all enjoy specific food, but we keep trying new ones because there might be something new there.
Predicting scheduling
Previous studies specialist Tudor Dragoi, then a graduate student at Boston University, led the study in which he and other members of the Sur Lab explored how humans and marmosets, a little primate, make predictions about event timing.
A straightforward task was assigned to three humans and two marmosets. They simply had to hit a button when the image disappeared ( marmosets poked a tablet while humans clicked a mouse ) and the image would appear on a screen for a while ( the amount of time varied from one trial to the next within a narrow range ).
The key to success was to avoid hitting the “funny” button too soon when a picture disappeared as quickly as possible. On successful trials, marshets were given a juice reward.
Marmosets required more training than humans, but the subjects all settled into the same acceptable pattern of behavior for the task. The longer the image stayed on the screen, the faster their reaction time to its disappearance.
This behavior is in line with the “hazard model” of prediction, which states that the longer an image is still present, the more likely it is to vanish very quickly. The subjects learned this and overall, with more experience, their reaction times became faster.
But as the experiment continued, Sur and Dragoi’s team noticed something surprising was also going on. Even though they had already learned what to do, mathematical modeling of the reaction time data revealed that both the humans and the marmosets were letting the outcomes of the immediate prior trial influence what they did on the following trial.
Subjects would decrease reaction time slightly ( presumably anticipating a shorter image duration once more ) while subjects would increase reaction time if the image remained on the screen for a while ( presumably because they figured they’d have a longer wait ).
Those results add to ones from , a similar study  , Sur’s lab published in 2023 in which they found that even after mice learned the rules of a different cognitive task, they’d arbitrarily deviate from the winning strategy every so often.
In that study, like this one, learning the successful strategy didn’t prevent subjects from continuing to test alternatives, even if it meant sacrificing reward.
The persistence of behavioral changes even after task learning may be a result of exploration as a method for identifying and building an optimal internal model of the environment, according to the scientists in the new study.
Relevance for autism
Sur said that the similarities between human and marmoset behaviors are also significant. That’s because , differences in making predictions , about one’s environment is posited to be a salient characteristic of autism spectrum disorders.
Some labs have begun work to create marmoset autism models because marmosets are small, social, and have more cognitive development than mice, but a key factor was establishing that they effectively model autism-related behaviors.
The study also adds weight to the growing notion that marmosets can indeed provide informative models for autism studies by showing that marmosets model neurotypical human behavior in terms of predictions.
In addition to Dragoi and Sur, other authors of the paper are Hiroki Sugihara, Nhat Le, Elie Adam, Jitendra Sharma, Guoping Feng and Robert Desimone.
Funding: The Simons Foundation Autism Research Initiative supported the research through the , Simons Center for the Social Brain , at MIT.
About this news about neuroscience research
Author: David Orenstein
Source: Picower Institute at MIT
Contact: David Orenstein – Picower Institute at MIT
Image: The image is credited to Neuroscience News
Original Research: Open access.
Mriganka Sur and colleagues ‘” Global to local influences on temporal expectation in marsets and humans” are available online. Current Biology
Abstract
Global to local influences on marmosets and humans ‘ temporal expectations
Marmosets are rapidly emerging as experimental models for studying the neural foundations of cognition and, crucially, for modeling human cognition disorders, but many aspects of their mental characteristics are still undefined.
Humans implicitly use prior information to predict upcoming events and lessen perceptual and decision-making uncertainty when judging elapsed time. The hazard rate model, which posits the likelihood of an event occurring in the future, provided it has not already occurred, is a powerful model of temporal expectation.
We report that, in accordance with the global task structure, marmosets trained on a reaction time task acquire the hazard rate model of expectation.
The model gradually evolves as it progresses with practice, but unanticipated local contingencies continue to influence it, as evidenced by a serial effect on responses ‘ response duration.
A multiple regression model and Bayesian updating of the hazard function both well describe the combined effects of global and local task structure and their computational implications.
Similar results from parallel human experimentation show that local influences influence reaction times and temporal expectation, as well as global influences.
Thus, task history and local structure in both marsets and humans constantly update task-specific responses, surprisingly at the expense of optimal responses after competent acquisition of an internal model.
