Neuroscientists decode how people juggle multiple items in working memory

When we try to remember more than one thing at a time, our brains don’t treat each item equally. New research published in Science Advances shows that the brain dynamically adjusts how much effort it devotes to each item based on its importance. By tracking brain activity while people remembered two locations on a screen, scientists discovered that two brain regions—the visual cortex and the frontal cortex—work together to allocate more resources to items deemed more important. This results in sharper, more precise memory for high-priority items, while low-priority ones are remembered with less clarity.

The research team, led by Hsin-Hung Li at The Ohio State University, wanted to understand how the brain deals with the fact that working memory is limited. Working memory helps us temporarily store information we need for things like making decisions, solving problems, or navigating the world. But it can only hold a small amount of information at a time. Past studies have shown that people can improve their memory for important items by prioritizing them. But scientists did not know how the brain manages this internally or what parts of the brain make those decisions.

To answer this, the researchers developed a method to measure how the brain represents multiple items held in memory at once and how it adjusts those representations based on priority. They combined functional MRI brain scans with a sophisticated computational model that could decode what participants were remembering—and how precisely they were remembering it.

The experiment involved 11 participants who each took part in several scanning sessions. While lying in an MRI scanner, participants looked at a screen where two colored lines appeared briefly—each in a different half of the screen. Before the lines appeared, a cue indicated which side of the screen was more likely to be tested later, marking one of the items as high-priority and the other as low-priority. After the lines disappeared, participants had to wait 12 seconds while trying to remember their locations. Then, they were prompted to look with their eyes at the location of one of the items, using a quick eye movement known as a saccade.

In most trials, the high-priority item was the one tested, but in a smaller number of trials, the low-priority item was selected instead. Eye-tracking data allowed researchers to measure how accurate each memory was, depending on whether it was high or low in priority.

Free daily newsletter

The researchers found that participants were consistently better at remembering the high-priority item. Their eye movements landed closer to the correct location, and they responded more quickly when recalling high-priority information. These results confirmed that people can adjust their memory performance based on what they think will matter most.

What made this study unique was its ability to decode brain activity related to each memory item—separately and in real time. Using a model rooted in a theory called probabilistic population coding, the team translated patterns of brain activity in the visual cortex into predictions about what the participant was remembering and how certain their brain was about it. They showed that high-priority items were encoded with more “gain,” meaning the brain devoted more neural resources to representing them. These representations were not only stronger but also more precise.

By analyzing data from both the visual and frontal areas of the brain, the researchers were able to go a step further. They found that activity in the frontal cortex predicted how much emphasis was placed on the high-priority item. The frontal cortex didn’t store the memory itself, but it seemed to decide which item deserved more attention. Once that decision was made, it influenced how the visual cortex represented each item. This suggests a kind of top-down control: the frontal cortex acts like a manager, telling the visual system how to distribute limited memory resources.

In trials where the high-priority item received more gain in the visual cortex, participants made fewer mistakes when recalling it. The researchers also showed that when the brain was less certain about an item—reflected by more “noisy” neural activity—participants took longer to respond and were more likely to make errors. This indicated a strong link between the brain’s internal sense of certainty and actual behavior.

The study also identified a specific part of the frontal cortex, the superior precentral sulcus, as playing a key role in prioritization. Activity in this region increased when participants were allocating more memory resources to the high-priority item. Its activity also predicted quicker response times, especially when participants felt more certain about the remembered location. In addition, the researchers found involvement from parts of the parietal cortex and a region in the temporal lobe called the posterior inferior temporal cortex—areas that are known to participate in attention and visual processing.

While the results help clarify how different brain regions coordinate to prioritize memories, the study had some limitations. The sample size was relatively small, with only 11 participants, which may limit how generalizable the findings are. The tasks were also highly controlled and may not reflect how working memory operates in everyday settings where people remember many types of information at once, not just spatial locations.

Another limitation is that although the researchers could estimate how much “gain” was assigned to each item, they could not directly measure how this translated into specific neural firing rates at the level of individual neurons. Future studies might combine brain imaging with methods that can capture more detailed neural activity to get a fuller picture of how memory prioritization works.

Even with these caveats, the study offers a new way to look at working memory. It suggests that rather than being a fixed bucket that can only hold a few items, working memory is a flexible system that can shift resources from one item to another depending on its relevance. The frontal cortex helps manage this process by signaling which memories deserve more attention, while the visual cortex adjusts how clearly those memories are encoded.

Lead author Hsin-Hung Li noted that this is one of the first studies to decode the brain’s memory for two different items on a single trial. He believes this technique could be useful for exploring many situations where people try to hold multiple thoughts in their minds at once. “There are so many situations in which people are trying to hold multiple thoughts in their minds and it is very useful to be able decode more than one,” he said.

The study, “Neural mechanisms of resource allocation in working memory,” Hsin- Hung Li, Thomas C. Sprague, Aspen H. Yoo, Wei Ji Ma, and Clayton E. Curtis.