Summary: New research uncovers how the brain processes learning by identifying the exact moment an animal learns a new skill. By observing individual neurons in mice, scientists found that learning occurs much faster than previously thought, in as few as 20 to 40 tries.
Surprisingly, this learning activity takes place in the sensory cortex, a region typically associated with perception rather than cognition. The study also found that even after mastering a task, mice continued to make errors, suggesting they were deliberately testing their knowledge.
Key Facts:
- Rapid Learning: Mice learned tasks in just 20-40 attempts, much faster than expected.
- Sensory Cortex Role: Learning occurred in the sensory cortex, not just higher-order brain areas.
- Error Testing: Mice continued making mistakes after learning, suggesting strategic exploration.
Source: Johns Hopkins University
By revealing for the first time what happens in the brain when an animal makes a mistake, Johns Hopkins University researchers are shedding light on the holy grail of neuroscience: the mechanics of how we learn.
The team pinpointed the exact moment mice learned a new skill by observing the activity of individual neurons, confirming earlier work that suggested animals are fast learners that purposely test the boundaries of new knowledge.
The federally funded work, which upends assumptions about the speed of learning and the role of the sensory cortex, and which the researchers believe will hold true across animal species including humans, is newly published in Nature.
“Looking at a tiny part of the brain in a mouse, we can understand how the brain learns, and we can makes predictions about how the human brain might work,” said Kishore Kuchibhotla, a Johns Hopkins neuroscientist who studies learning in humans and animals.
“The field of neuroscience has made great progress decoding motor activity and how the brain processes sight and sound. But a holy grail of this type of research is thought—what comes between the hearing and the doing—we’re all still trying to understand the patterns of brain activity that underly higher-order cognitive processes. These findings are a step in that direction.”
Although the ability to learn quickly would benefit any animal in the wild, animals studied in labs seem to learn slowly and methodically. It typically takes mice, for instance, thousands of tries to learn a task, several hundred at best.
Kuchibhotla’s lab previously found that animals’ performance doesn’t necessarily sync with their knowledge—or that animals might know a lot more than they demonstrate in tests. The lab also found that animals that seem to be slow learners might be testing their new knowledge. But by merely watching animals struggle at tasks, they couldn’t tell a slow learner from a strategic tester of boundaries.
“We are interested in the idea that humans and other animals may know things about the world, things that they choose not to show,” Kuchibhotla said. “Our core question is what is the neural basis of this distinction between learning and performance.”
The researchers taught mice to lick when they heard one tone but not to lick when they heard a different sound. From the moment training began, the team recorded the activity of neurons in the auditory cortex, an area of the brain associated with hearing and perception.
There were two major surprises. First, the mice learned in 20 to 40 tries, “extraordinarily fast,” according to Kuchibhotla. And second, this learning activity happened in the sensory cortex, something that has typically been associated with nonsensory brain areas.
“This work illustrates the importance of assessing how brain activity impacts behavior at different stages of the learning process and in different conditions,” said first author Celine Drieu, a postdoctoral fellow studying neuroscience at Johns Hopkins. “Our results show that a sensory cortex does more than processing sensory inputs; it is also crucial to form associations between sensory cues and reinforced actions.”
When the mice continued to make errors, licking at the wrong times long after their neural activity showed they’d learned the task, their brain activity confirmed to the researchers that the mice knew the rules of the game—they were just experimenting.
‘We were able to decode the cognitive driver of an error,” Kuchibhotla said. “We could tell if the animal was making a mistake or just wanted to give the other option a shot.”
Once the mice had mastered the task and ceased their exploratory behavior, this higher-order activity started to diminish, and the sensory cortex was no longer involved in the task.
“We think this means that animals are smarter than we think, and that there are distinct brain dynamics related to learning. You might know something, but there’s a parallel process related to how you use it. The brain seems wired to do that well, to allow us to toggle between performance and learning as we get better and better at something.”
About this learning and neuroscience research news
Author: Jill Rosen
Source: Johns Hopkins University
Contact: Jill Rosen – Johns Hopkins University
Image: The image is credited to Neuroscience News
Original Research: Closed access.
“Rapid emergence of latent knowledge in the sensory cortex drives learning” by Kishore Kuchibhotla et al. Nature
Abstract
Rapid emergence of latent knowledge in the sensory cortex drives learning
Rapid learning confers significant advantages on animals in ecological environments. Despite the need for speed, animals appear to only slowly learn to associate rewarded actions with predictive cues.
This slow learning is thought to be supported by gradual changes to cue representation in the sensory cortex. However, evidence is growing that animals learn more rapidly than classical performance measures suggest, challenging the prevailing model of sensory cortical plasticity.
Here we investigated the relationship between learning and sensory cortical representations. We trained mice on an auditory go/no-go task that dissociated the rapid acquisition of task contingencies (learning) from its slower expression (performance).
Optogenetic silencing demonstrated that the auditory cortex drives both rapid learning and slower performance gains but becomes dispensable once mice achieve ‘expert’ performance. Instead of enhanced cue representations, two-photon calcium imaging of auditory cortical neurons throughout learning revealed two higher-order signals that were causal to learning and performance.
A reward-prediction signal emerged rapidly within tens of trials, was present after action-related errors early in training, and faded in expert mice. Silencing at the time of this signal impaired rapid learning, suggesting that it serves an associative role.
A distinct cell ensemble encoded and controlled licking suppression that drove slower performance improvements. These ensembles were spatially clustered but uncoupled from sensory representations, indicating higher-order functional segregation within auditory cortex.
Our results reveal that the sensory cortex manifests higher-order computations that separably drive rapid learning and slower performance improvements, reshaping our understanding of the fundamental role of the sensory cortex.
