When the Brain Directs Attention Toward Its Own Map
New Research from the Kavli Institute at NTNU
It is not every day that Norwegian research is published in Science. The journal is among the most prestigious scientific publications in the world. Competition to pass through its narrow gate is fierce, and the articles that succeed are expected not only to be scientifically rigorous. They must also move their field of research a significant step forward.
There is therefore good reason to pause and consider a new study from the Kavli Institute for Systems Neuroscience at the Norwegian University of Science and Technology, NTNU, in Trondheim. The study was conducted by Abraham Z. Vollan, Michael F. Schellenberger, Richard J. Gardner, May-Britt Moser, and Edvard I. Moser. It was published in Science on 6 August 2026 under the title “Adaptive Modulation of Theta Sweeps in the Brain’s Navigation Circuit.”
The title is not immediately accessible to everyone. It tells us that the researchers have investigated how so-called theta sweeps change within the brain’s navigation circuit. Behind this technical formulation, however, lies a question that is also of interest to practical philosophy:
How does the brain find its way through the world?
Not only in the sense of: How do we know where we are? But also: How does the brain orient itself toward what matters at this particular moment? How does it select a place, a goal, or a possible path from among all the possibilities that surround us?
The new research from Trondheim offers us a fascinating glimpse into this process.
The Inner Map
May-Britt Moser and Edvard Moser are known for their discovery of grid cells in the brain’s entorhinal cortex. Together with place cells in the hippocampus, these cells form part of what is often described as the brain’s navigation system.
Place cells become active when an animal is in a particular location. One cell may be especially active in one area of a room, while another cell is activated somewhere else. Grid cells form a more regular pattern. Their activity can be compared to an internal coordinate system that makes it possible to calculate distances, directions, and positions.
In other words, the brain does not merely possess a collection of isolated memories of places. It creates an inner map.
Such a map is not, however, a map in the ordinary sense. It does not lie ready and folded inside the brain like a road map in the glove compartment of a car. It is created and recreated through activity in large networks of neurons. The map is alive. It changes as the animal moves, senses, searches, and acts.
But how does the brain use this map?
The new study shows that the map does not merely tell the animal where it is. The brain also investigates places ahead of the animal—and sometimes places behind it. It can direct the inner map toward a target before the body has begun to move toward it.
The Brain Investigates Possible Paths
Activity in the hippocampus and entorhinal cortex is organised by a rhythm known as the theta rhythm. In a moving rat, this rhythm usually oscillates approximately eight times per second.
Within each of these brief theta cycles, groups of neurons are activated in rapid sequences. The sequences represent not only the place where the rat currently is. They can also represent a possible path through the space ahead of it.
The researchers call these sequences theta sweeps.
A theta sweep can be understood as a brief investigation of a possible route through the inner map. Within a fraction of a second, neural activity moves from the representation of the animal’s present location toward places it has not yet reached.
When a rat explores an open area without a specific goal, these sweeps usually alternate between the left and right sides. One sweep investigates the space slightly to the left. The next investigates the space slightly to the right. In this way, the brain covers a broad area in front of the animal.
It resembles a searchlight moving rhythmically from side to side in the darkness.
The brain is therefore not content merely to register its surroundings passively. It investigates them. It scans possibilities.
But what happens when one particular possibility becomes more important than all the others?
The Rat Pursuing a Moving Target
Using Neuropixels probes, the researchers recorded activity from several hundred neurons simultaneously. These probes made it possible to follow activity in grid cells, place cells, and direction-sensitive cells while the rats moved freely.
First, the rats searched for small pieces of food scattered randomly across a large, open arena. They were then given a very different task. They had to chase a piece of food suspended from a string and moved rapidly and unpredictably around the arena by an experimenter.
The task resembled a simple form of hunting. The rat had to follow the target continuously and change direction whenever the food moved.
During free exploration, the theta sweeps continued to alternate broadly between left and right. The rat’s brain investigated many possible directions.
During the pursuit, the pattern changed.
The sweeps became narrower and gathered around the direction of the food. At the same time, the theta rhythm accelerated. The brain therefore investigated a smaller area, but it investigated this area more frequently and more intensely.
The searchlight ceased to move broadly across the landscape. It concentrated its beam on the target.
This is the study’s first main finding: The brain’s inner map can rapidly reorganise itself according to what matters at that particular moment.
This happens even though the target is not located at a previously learned place. The food is moving constantly. The rat cannot simply retrieve a familiar route or return to a place where it has previously received a reward. The navigation system must adapt from moment to moment.
The Map Turns Before the Body
One of the most interesting findings emerged when the food suddenly changed direction.
The researchers could then see that activity in the inner navigation system turned toward the food’s new position before the rat began to turn its head. On average, the neural reorientation occurred approximately 130 milliseconds before the head movement.
One hundred and thirty milliseconds is a very short time. Yet in the work of the brain, it is long enough for something important to happen.
The inner map has already oriented itself toward the target while the body is still facing another direction.
This does not necessarily mean that theta sweeps determine what the rat will do. The researchers are careful on this point. The same event—the movement of the food—may initiate both the neural reorientation and the subsequent head movement. The study therefore does not show that the sweep alone causes the movement.
It does show, however, that the brain’s inner orientation is not bound to the actual direction of the head. The map can temporarily free itself from the body’s orientation and turn toward another place.
Sometimes the neural activity continued to follow the food even when it moved behind the rat. In some cases, the rat remained still while the theta sweeps continued to track the target. The same thing occurred when the experiment was conducted in complete darkness.
The activity therefore cannot be explained simply as a reaction to what the rat sees. Nor can it be reduced to a completed command to move the head or body in a particular direction.
The brain appears to hold the target within its inner map.
When the Rat Stands Still
The navigation system continued to work even when the rats stopped moving.
During movement, the theta sweeps were organised relatively regularly around the direction of the head. During immobility, the relationship between head direction and the internal direction signal became looser. The sweeps could turn toward sounds, incoming pieces of food, or other events in the environment.
Some of these neural turns were followed by the rat turning its head or beginning to move. Others were not.
This is important. The inner map does not represent only movements that are actually carried out. It can also investigate places without the investigation ending in visible action.
Not every possible path becomes an actual path.
We recognise something of this in our own lives, although we clearly cannot transfer findings from rats directly to human beings. We can direct our attention toward a possibility without pursuing it. We can consider a path and decide not to take it. We can turn toward something in thought while the body remains still.
The study does not show that the rat considers its possibilities in the same conscious way as a human being. But it does reveal a neurobiological system capable of orienting itself toward possible places and actions before any outward movement takes place.
Moving Backwards
The researchers also conducted another simple and elegant experiment.
After eating a reward at the far end of a narrow, dead-end corridor, the rat had to back out. Its head pointed forward, while its body moved backwards.
What would the inner navigation system follow: the direction of the head or the direction of movement?
The classical head-direction cells continued to represent the direction in which the head was pointing. This signal remained stable. The theta sweeps, however, reversed and followed the backwards direction of movement.
The brain therefore maintained two different spatial signals at the same time. One preserved the body’s orientation in space. The other directed the inner investigation toward the area into which the rat was actually moving.
This shows that theta sweeps are not merely an automatic extension of gaze or head direction. They can be adapted to the action being performed.
The navigation system does not ask only: Which way am I facing?
It also asks: Where am I going?
The Map Also Works During Sleep
The findings become even more remarkable when the researchers examine brain activity during rapid eye movement, or REM, sleep.
The rat lies still. It is not moving through the arena. Sensory input is strongly limited. Yet grid cells and direction cells continue to form organised patterns.
During REM sleep, the researchers identified periods in which the theta rhythm suddenly accelerated. When this happened, the inner sweeps became faster and more concentrated, much as they did when the awake rat followed a moving target.
The navigation system could therefore move through the inner map without the rat moving through external space.
We should be careful about calling this dreaming. The researchers cannot know what the rat may be experiencing. They record electrical activity, not the content of a dream. Nevertheless, the finding shows that the brain can produce the same fundamental dynamics without direct guidance from the senses or from outward action.
The inner map can work from within.
The Parasubiculum—A Possible Control Room
The researchers draw particular attention to a region known as the parasubiculum. This area lies close to the hippocampus and entorhinal cortex and receives information from the brain’s classical direction system.
The classical head-direction cells function somewhat like a stable inner compass. They continue to indicate the direction in which the head is pointing, even when the theta sweeps turn toward a target at the side or follow a backwards movement.
In the parasubiculum, this stable compass signal appears to be combined with other information: information about the target, the movement, and what matters in the immediate situation. Here, the internal direction can become separated from head direction and guide the theta sweeps toward another place.
The researchers therefore describe the parasubiculum as a possible control point for the flexible direction of the brain’s spatial representations.
The compass is not lost. It continues to keep track of direction. But another system can use the compass to direct attention where the organism needs it.
Attention Within the Inner Map
The researchers compare theta sweeps to active sensing.
When we look for something, we move our eyes. When a rat investigates its surroundings, it moves its whiskers. A bat does not emit its echolocation signals randomly but directs them toward those parts of its surroundings that matter. Sensing is not merely something that happens to an organism. The organism actively participates in how the world is sensed.
Theta sweeps may be an internal version of the same principle.
Here, the brain does not direct an eye, an ear, or a whisker toward the world. It directs its activity toward particular parts of its own map of the world. It selects places for more intensive processing.
The researchers therefore use the expression attention-like mechanism.
This is a precise and important qualification. They do not claim to have discovered attention as such. Nor do they claim that the rats are necessarily consciously attending to the possible paths that are represented. Theta sweeps do, however, possess some of the fundamental characteristics of attention: They select something, give it priority, and allow it to occupy a greater place in subsequent processing.
During free exploration, many possibilities are kept open. When a target becomes important, the field narrows.
The Brain Does Not Merely Depict the World
It is here that neuroscience meets practical philosophy.
We can easily imagine sensation and knowledge as a form of representation. The world is outside us. The senses receive information. The brain processes that information and creates an inner representation of what exists out there.
But this study reveals something more dynamic.
The brain does not merely form an image of the world. It directs itself toward the world. It selects something as significant. It investigates some possibilities more thoroughly than others. The inner map is shaped by what the organism is doing.
The relevant place is not necessarily the place where the animal currently is. It may be the place where the food is, the place toward which movement is leading, or the place that might be explored.
Knowing is therefore not detached from acting. What becomes clear within the map is related to what the organism needs to do.
This does not mean that the world is merely a construction of the brain. The food exists out there and moves independently of the rat. The corridor has a real direction. Obstacles and distances do not disappear because the brain perceives them in a particular way.
But neither does the world appear with the same significance everywhere and at all times. Something comes to the foreground. Something else recedes. What matters now occupies more space than what might have mattered under different circumstances.
This is also true of human life. We do not live in a world where everything is equally significant at every moment. Attention gathers the world around something. A task. A person. A danger. A hope. A path we are trying to find.
What the Research Shows—and What It Does Not Show
The study involved ten rats, while some of the individual experiments included fewer animals. We therefore cannot simply draw conclusions about human consciousness, free will, imagination, or decision-making.
Nor have the researchers shown that theta sweeps alone produce action. When a sweep turns toward a target before the rat moves its head, both the brain activity and the movement may be responses to the same external event.
What the researchers have shown is nevertheless significant:
The brain’s navigation system is not a passive map of the animal’s position. It is a flexible system capable of continuously reallocating its activity toward the places that matter in the ongoing situation.
This adaptation takes place within fractions of a second. It can be observed during targeted pursuit, free exploration, backward movement, immobility, and, in another form, during REM sleep. The sweeps can be influenced both by external events and by internally generated patterns of activity.
It is the breadth of these findings that makes the study so interesting.
Honour and Recognition for the Kavli Institute
There is good reason to give honour and recognition to the research community at the Kavli Institute for Systems Neuroscience at NTNU.
Over many years, this community has built a research tradition that combines bold questions with exceptional methodological precision. The discovery of grid cells changed our understanding of how the brain represents space. The new study carries this work forward by showing how the spatial system is used flexibly from moment to moment.
It is not often that research conducted in Norway is given such prominence in Science. When this happens, it is not merely a recognition of the individual article. It is also a recognition of a research community that, over a long period, has developed ideas, methods, and researchers at the highest international level.
At the same time, it is worth noting that this is basic research. The researchers are not primarily trying to develop a particular product or an immediate treatment. They are trying to understand how the brain finds its way through the world.
This is precisely why the research may become important.
Basic research often begins with a question that may appear simple: How does an animal know where it is? But when the question is examined thoroughly, it opens toward memory, attention, action, planning, and sleep. Ultimately, it also touches upon the question of how a living organism creates coherence between itself and its surroundings.
The Long Road from the Rat Brain to the Patient
At the same time, it is necessary to emphasise how far this research is from the diagnosis and treatment of human beings.
The experiments were conducted on healthy, young rats. The researchers investigated fundamental mechanisms in the brain’s navigation system. They did not study Alzheimer’s disease or any other neurological disorder, nor did they develop a diagnostic method or test a treatment.
The road from “we understand the mechanism better” to “we can help a patient” is a very long one.
Before such findings can acquire any clinical significance, researchers must determine whether corresponding patterns of activity exist in human beings, whether they can be recorded reliably without electrodes placed inside the brain, and whether they change in a recognisable way in particular diseases. They must then show that such changes can distinguish disease from normal variation and ageing, and that they can contribute to more reliable diagnosis, assessment of disease progression, or treatment.
Each of these stages may take many years. There is no guarantee that the road will lead to a clinical application.
The Neuropixels probes used in this study record activity directly from hundreds of neurons in the rat brain. Such a method cannot simply be applied to human beings. In human research, scientists generally have to look for more indirect traces of the same mechanisms, for example through electroencephalography, or EEG, and magnetoencephalography, or MEG. Such measurements may eventually contribute to improved biomarkers—measurable indicators of how particular networks in the brain function or change.
A more immediate practical consequence of this kind of basic research is therefore not a new treatment, but better questions, better animal models, and more precise interpretations of brain activity.
The knowledge may, for example, be used to investigate whether diseases affecting the hippocampus and entorhinal cortex, including Alzheimer’s disease, also disrupt the brain’s rhythmic organisation of space, memory, and attention. It may help researchers improve animal models of the disease or search for possible EEG- and MEG-based biomarkers (National Institute on Aging, 2024; Sohrabpour et al., 2025).
But this, too, is a long and uncertain road. A possible biomarker must undergo extensive investigation before it can find any place in clinical practice. Researchers must show that the measurement is stable, that it is genuinely related to the disease, and that it provides knowledge that is useful for the individual patient. Many promising findings from animal studies prove difficult to reproduce in human beings. Others can be reproduced but are not sufficiently precise for diagnosis or treatment.
This does not make basic research less significant. On the contrary, it shows why basic research requires both patience and restraint. It does not necessarily give us a treatment. First, it gives us a better understanding of what it is we are trying to treat.
Honour and recognition for the Kavli Institute should therefore not be based on promises of an imminent medical benefit that the study does not support. The achievement lies in the researchers having uncovered something new about how a living brain orients itself in the world.
Where this knowledge may one day lead, we do not yet know.
The Map and the Road
A map is not the same as the world it depicts. Nor is it the same as the road that is actually travelled.
The new research from the Kavli Institute nevertheless shows that the brain’s map already contains a movement toward the world. It does not wait passively for the body to arrive. It searches, investigates, and directs itself toward possible places.
Some of these possibilities become actions. Others disappear again.
Perhaps this is how we should understand the living map: not as a finished drawing of the world, but as a continually unfolding conversation between the place where we are, that which calls for our attention, and the road not yet travelled.
References
National Institute on Aging. (2024, October 21). Clinical interventions and diagnostics portfolio. https://www.nia.nih.gov/research/dn/clinical-interventions-and-diagnostics-portfolio
Sohrabpour, A., Sarovic, D., Zetterberg, H., Hämäläinen, M. S., Baillet, S., & Khan, S. (2025). Tracking electrophysiological signatures of Alzheimer’s disease: A systematic review of multimodal studies. Alzheimer’s & Dementia, 21(11), e70835. https://doi.org/10.1002/alz.70835
Vollan, A. Z., Schellenberger, M. F., Gardner, R. J., Moser, M.-B., & Moser, E. I. (2026). Adaptive modulation of theta sweeps in the brain’s navigation circuit. Science, 393, eaef4184. https://doi.org/10.1126/science.aef4184
This essay was written in a conversation with Claude/Anthropic and OpenAI/ChetGPT