Using a classic visual illusion, 91��ɫ researchers identify a history-dependent computation present in primate vision that current AI models fail to reproduce.

TORONTO, Oct. 2, 2026 – Our eyes do not always tell us exactly where things are – and that may be a feature of how biological vision works, rather than simply a flaw. A new study by 91��ɫ researchers uses a common illusion to ask if artificial intelligence is meant to see more like us, should it make some of the same systematic perceptual “mistakes”?

For example, after staring at something moving steadily in one direction, a stationary object viewed immediately afterward can appear slightly displaced in the opposite direction. This well-known visual illusion, called a motion aftereffect, gives scientists an unusual window into the computations underlying perception: the image itself has not moved, but our experience of where it is has changed.

“Today’s AI vision systems are impressive, but they still do not always see the world the way we do. This study captures the promise of NeuroAI and what it can do when neuroscience and artificial intelligence are brought together. By using smart experiments to reveal the computations biological vision uses and AI still lacks, we can use those insights to build better, more brain-like artificial systems,” says senior author 91��ɫ Assistant Professor Kohitij Kar, the Canada Research Chair in Visual Neuroscience and a member of 91��ɫ’s Centre for Vision Research and Centre for Integrative and Applied Neuroscience.

Current AI vision systems can often determine where an object is accurately, but they generally do not reproduce the way recent visual experience can reshape that answer. In humans, staring at motion can make a subsequently viewed stationary object appear displaced even though its pixels have not moved. The researchers found a corresponding change in the primate visual cortex – but not in the AI models they tested.

To gain better insight into the issue, the researchers examined whether artificial neural network models capture the same history-dependent changes in spatial representations seen in biological vision, or whether their position representations primarily reflect the physical properties of the image. The researchers, including the paper’s first author and 91��ɫ graduate student Elizaveta Yakubovskaya, used precise measurements to find out where AI differs from biological vision and how that gap could be bridged.

“Combining recordings from primate visual cortex with human perception experiments, we used motion adaptation to induce a visual illusion and make a stationary object appear slightly shifted in position, then asked whether the brain and AI showed the same effect. Human observers reported the illusion, and neural representations of position in the primate inferior temporal (IT) cortex shifted in the same direction, even though the image itself had not changed,” says Yakubovskaya.

The researchers leveraged motion adaptation to show where perceived and pixel-based positions diverge, allowing them to test the behavioral relevance of IT codes. The findings not only further our understanding of IT’s role in spatial information encoding but provide a new benchmark to evaluate dynamic vision models.

“There is a growing question in AI about whether increasingly capable systems will become more like us or increasingly different from us,” says Kar of the Faculty of Science and a member of the 91��ɫ-led Connected Minds. “If we want AI that works with humans and understands the world in more human-compatible ways, we cannot focus only on whether it gets the right answer. We also need to understand the computations that produce human perception and behavior. Neuroscience gives us a way to discover those computations and, potentially, build them into AI,” adds Kar.

The study titled, , is published today in Current Biology.

The post Visual illusion reveals what today’s AI vision is missing – 91��ɫ study appeared first on News@91��ɫ.

The wiring and rewiring of the brain never ends. Neural pathways are constantly being reshaped as we interact with the world and learn new things. At 91��ɫ and MIT’s McGovern Institute, scientists are combining detailed analysis of brain activity with computational modelling to better understand that change.

91��ɫ Assistant Professor , McGovern Institute Postdoctoral Fellow , an affiliate member of Kar’s lab and a member of the Centre for Integrative and Applied Neuroscience at 91��ɫ, and Investigator  worked together to compare what happened when monkeys and an artificial neural network with brain-like architecture were trained to visually identify the same objects. As the model’s performance improved, it reorganized itself in ways that closely paralleled changes the team detected in the brains of monkeys.

Their research, reported  , shows how changes in visual processing support animals’ ability to learn to discriminate new kinds of objects. By modelling these changes, the researchers hope to better predict how training reshapes perception, which could one day inform educational strategies for a wide range of learners.

“Our visual brain does not undergo a drastic reconfiguration when we learn a new object. Instead, it subtly reshapes how visual information is represented, making the distinctions that matter easier for the rest of the brain to read out. What is powerful about this study is that artificial neural networks could predict these subtle brain changes, giving us a concrete bridge between neuroscience and AI,” says Kar. “By translating inferences from animal neuroscience into computational models that can generate human-testable predictions, this framework makes translation much more feasible. It allows us to move from understanding how learning changes the macaque visual system to perhaps asking in the near future: why some children may struggle to form stable, generalizable representations – such as recognizing letters across fonts, linking words to objects, distinguishing faces, or applying knowledge in new situations.”

Subtle changes

Learning about a new object calls on many parts of the brain. Visual-processing areas work together to make sense of information taken in through the eyes, then communicate with other brain areas to give the visual information meaning and guide behavior. Multiple parts of this system likely change during learning, and the research team wanted a clearer understanding of how that change is distributed.

Neuroscientists have debated how much change occurs in the brain’s visual-processing areas when an animal learns to recognize new objects. Some suspected that visual-processing pathways remain largely unchanged during learning to avoid broadly disrupting visual perception, but others have reported changes in activity within dedicated visual-processing areas with this kind of learning in humans and other primates. 

To take a closer look, the team focused on neural activity in a key component of the brain’s visual object-processing network, the inferior temporal (IT) cortex. By the time visual information reaches the IT cortex, key object features are clearly represented – so much so that it’s possible to “decode” what object a monkey is seeing and even predict what errors it’s likely to make in identifying it, simply by analyzing patterns of neural activity there.

The team recorded neural activity in the IT cortex from two groups of monkeys as the animals looked at and identified images of objects. Some of the monkeys were untrained, so the images they saw had little meaning to them. Others had already learned to identify similar objects, so they could usually discriminate between elephants, chairs and other select objects, even when those objects were presented at different sizes, from different angles, or against different backgrounds than the ones they had seen before.

The broad pattern of activity in the IT cortex was largely similar in trained and untrained monkeys, suggesting that learning had not dramatically rewritten this high-level visual representation. Still, the group found subtle but reliable differences in the way neurons in the IT cortex responded to images in monkeys that had learned to recognize the kinds of objects they were shown, compared to the untrained monkeys.

Modelling learning

The group turned to computational models to investigate how those modest changes might contribute to learning. Sörensen trained a suite of artificial neural networks whose internal components had been mapped to monkey IT cortex to identify the same categories of objects the monkeys had seen. The models were designed to learn using gradient descent, meaning they continually improved their accuracy by adjusting their parameters in response to errors.

Only some of the primate-like models showed learning behavior that matched that of the monkeys. In those that did, the IT-like stage changed in ways that resembled the learning-related changes the researchers had observed in the IT cortex of trained monkeys.

While gradient descent is commonly used to train artificial intelligence, it is generally considered biologically implausible as a direct model of how the brain learns. The researchers say the strong match in learning effects between the monkeys and their model demonstrates that these kinds of artificial neural networks can offer insights into biological learning at a useful level of abstraction, even if the brain does not learn in the same way.

“This shows that you can actually build in silico versions of future experiments,” Sörensen says. “I think that gives us this playground of asking ‘what if’ questions – and potentially predicting new things that go beyond the experimenter’s intuition.”

Most of the changes that allowed for learning in the model occurred outside of the IT cortex. “This tells us that there is a lot between the area we recorded from and the final behavioral readout that needs to change during this process,” Kar says. He adds that the team’s model will be useful as researchers look more deeply into how downstream brain areas contribute to learning.

The researchers stress that their study allowed more granular measurements of brain activity than would be possible in humans, and because monkeys’ brains are organized similarly to our own, their experiments have direct relevance to human learning. They say understanding the impact of plasticity in monkeys’ IT cortex could help researchers design new learning strategies for humans.

“Our prior conceptual working model of you – or a monkey – learning new objects was that your brain makes changes to synaptic connections that are largely downstream of your visual system, so you don’t destroy your visual system,” says DiCarlo, who is also the Peter de Florez Professor of Brain and Cognitive Sciences and director of the MIT Siegel Family Quest for Intelligence. “You wouldn’t want your whole visual system to become an elephant detector [just because you’ve learned to identify an elephant]. But this study went beyond that to say actually, when you learn ‘elephant,’ your IT does change a little bit to make it a little more relevant to elephants.”

That likely has consequences for recognizing other visual features, too. Subtle changes in the IT cortex that support elephant recognition might also make you better at identifying things other than elephants, DiCarlo says. Likewise, the same changes might make it a little harder to identify something else.

These kinds of consequences may be difficult to predict intuitively, but become obvious with computational modelling. For instance, the team’s models revealed that after learning to recognize new objects, the IT cortex contained more information about objects’ locations. By providing insights like these, models could aid the design of more effective training strategies for visual tasks, including for people with altered sensory processing, who may learn from visual information in atypical ways.

With files from MIT's McGovern Institute

The post Learning to identify new objects reshapes parts of the brain, research finds appeared first on News@91��ɫ.

A new study out of 91��ɫ shows autistic individuals more likely to use both hands

TORONTO, Jan. 19, 2026 – Most people take for granted which hand they use to reach for a cup of coffee or a puzzle piece. However, a new study out of 91��ɫ suggests that for autistic individuals, which hand they use for various tasks is highly variable, which points to profound differences in the brain.

The research, published today in the journal Autism Research, found that even autistic adults who are right-handed demonstrate a reduced specialization of hand use and more distinctive movement patterns when compared to non-autistic peers.

“Handedness is one of the most visible markers of how our brain's left and right hemispheres specialize for different tasks,” says 91��ɫ Associate Professor , who conducted the study with lead author and Master’s student Emily Fewster. “In the neurotypical population, about 90 per cent of people show a strong right-hand dominance, reflecting the left hemisphere’s specialization for fine motor skills. Our study shows that in autism, this specialization is less pronounced, leading to a unique and highly individualized motor signature.”

The LEGO Building Task

To observe these behaviours in a real-world context, researchers asked 54 right-handed adults, half with an autism diagnosis, to recreate complex LEGO models. Unlike traditional questionnaires that ask which hand someone uses to write, this naturalistic task allowed researchers to track thousands of dynamic movements in 3D space.

By analyzing how people actually move during the LEGO building task, researchers found that the right-handedness of autistic participants’ function quite differently than that of the non-autistic participants. Despite both groups identifying as right-handed, the autistic participants used their right hand much less often for grasping and did not show the typical dominant preference for using their right hand when reaching across their body.

The autistic participants also tended to shrink their workspace by focusing on blocks placed closer to them, suggesting a more cautious or individualized strategy for managing the space around them. In addition, their movements followed highly unique, idiosyncratic paths. While non-autistic participants tended to follow a similar sequence of actions, each autistic participant moved in a distinct, more variable way.

Together, these findings suggest that the autistic brain organizes movement in a less specialized, more variable manner than previously understood.

Implications for Earlier Identification

While the study focuses on brain organization, these "motor signatures" have significant clinical potential. Because motor skills often emerge in infancy, long before the complex communication skills typically used to diagnose autism, identifying these subtle motor differences could open a window for much earlier support.

“Standard questionnaires often miss these nuances because they don’t capture the dynamic nature of real-life movement,” says Freud. “By looking at how people actually move in a natural setting, we can identify objective markers that might eventually help us provide more tailored support strategies much earlier in development.”

The researchers suggest that this "noisy" or variable motor processing supports the theory that autism involves broader, less specialized neural representations across the brain.

The post Understanding how right- or left-hand dominance could open a window into the autistic brain appeared first on News@91��ɫ.

TORONTO, Nov. 21, 2023 – Faced with images that break the expected pattern, like a do not enter sign where a stop sign is expected, how does the brain react and learn compared to being shown images which match what was predicted?

That was the question a team, including 91��ɫ, set out to answer. A long-standing theory suggests the brain learns a predictive model of the world and its internal predictions are updated when incoming sensory data proves them wrong. However, what the researchers found surprised them, says 91��ɫ Faculty of Science Associate Professor Joel Zylberberg, co-corresponding author of the newly published paper.

“Testing this theory has always been a challenge,” he says. “We needed to be able to measure the top-down signals to the sensory areas of the brain over long periods of time to show how the brain learns new sensory input patterns.”

Using a mouse model, the researchers displayed images of visual patterns over multiple days, then presented other images that violated those patterns, while measuring the brain’s activity in the visual cortex, where visual information from the retina is processed. The idea was to test how the neurons reacted to the new pattern-violating sensory information.

Several of the researchers, including Zylberberg, are Fellows in the ’s Learning in Machines and Brains group, which conducted the research as part of the Allen Institute for Brain Science’s Brain Observatory and its OpenScope program. OpenScope has been compared to an observatory where astronomers work together to study the universe, only this time researchers are sharing data to study the brain.

The measurements were taken at the neurons’ distal apical dendrites of the visual cortex, which receive top-down signals, and at their cell bodies, which receive bottom-up signals. They wanted to know if the distal apical dendrites processed visual stimuli differently from their cell bodies when the signals both matched and violated expected patterns.

It turns out, the brain’s response to image patterns that violate the brain’s predictions, evolves differently over time when compared to pattern-matching images.

“Surprisingly, the distal apical dendrites responses grew significantly over time becoming increasingly sensitive to inputs that violate the patterns, while the cell bodies lost their initially strong sensitivity,” says Zylberberg, a computational neuroscientist. “This finding could offer critical insight into sensory computation and predictive learning in the brain.”

The finding suggests that the pattern-violating stimuli drove the changes and different forms of pattern-violating stimuli may elicit different kinds of prediction errors than expected. It points to a component of the brain that could have a distinct and important role in sensory learning not previously known.

“Knowing how the brain processes new visual sensory information is important for developing better machine learning algorithms and applications which could hopefully help restore people’s sight in the future,” says Zylberberg.

The paper, , was published today in JNeurosci, the Journal of Neuroscience.

The post Researchers help unravel brain processes involved in vision appeared first on News@91��ɫ.

The post 91��ɫ U led Connected Minds explores how new technologies affect our brains, society and the most vulnerable appeared first on News@91��ɫ.

Together with Queen’s University, the cross-disciplinary work is backed by $105.7M in federal funding through the Canada First Research Excellence Fund

TORONTO, April 28, 2023 — Is an equitable world that includes humans and machines possible? 91��ɫ researchers believe it must be and have set out to make it so through a first of its kind interdisciplinary research initiative called Connected Minds: Neural and Machine Systems for a Healthy, Just Society.

From universities to industries, hospitals and policymakers, artists and Indigenous communities, 91��ɫ’s Connected Minds will engage 50+ community partners and research collaborators over seven years supported by a historic $318.4 million in funding. Connected Minds has received a combined , announced earlier today by the Government of Canada. Of that, 91��ɫ received $82.8 million and institutional partner Queen’s University received $22.8 million.

Led by 91��ɫ, Connected Minds brings together experts in multiple fields, including humanities, engineering, law, and life sciences, located across eight 91��ɫ Faculties and three Queen’s Faculties. Researchers will examine the ways in which emerging technology, like Artificial Intelligence, is transforming and entangling society – dubbed the ‘techno-social collective.’ Researchers will work to discover how to balance both the potential risks and benefits for humanity.  

Some of the program’s proposed projects include explorations into a more inclusive metaverse, virtual reality and community organizing, neurotechnologies for healthy aging, Indigenous data sovereignty, and how human brain function changes when people interact with AI versus each other.  

In addition to supporting interdisciplinary teams of researchers conducting research on promoting a healthy, resilient, and just techno-social collective, Connected Minds will fund 35 strategic faculty hires, partner-focused seed, team, and prototyping grants, knowledge mobilization and commercialization activities, and an ambitious multi-institutional micro-credential training program with 385 trainees and cross-sector stakeholders. All activities will require interdisciplinary participation, and projects that benefit Indigenous and other equity-deserving groups will be prioritized.

A key structural component of the program is an Indigenous-led focus and will feature a dedicated Indigenous research space on 91��ɫ’s Keele Campus, as well as employing an overarching decolonization, equity, diversity and inclusion (DEDI) strategy.

The program’s operations will involve a directorate led by internationally renowned neuroscientist Prof. , as Scientific Director, along with intellectual property and technology law expert Prof.  , as Vice-Director and Indigenous health scholar Prof. as Associate Director. Engineer and neuroscientist Prof. joins as the Vice-Director from Queen’s University.  

The governance structure of Connected Minds includes a Board of Directors to supervise the program ramp up, oversee its progress towards achieving strategic goals and maintain fiduciary responsibility, an External Advisory Board to advise on the overall strategic direction, knowledge mobilization and commercialization activities, and an Indigenous Advisory Circle to counsel all aspects of Indigenous engagement, including issues of privacy and data sovereignty.

The total value of the Connected Minds projects is $318.4 million with the remaining funds, including in-kind contributions, being contributed largely by multi-sector partners, municipal governments, and collaborating institutions.

QUOTES

“Receiving this second CFREF award in the last two competitions reflects 91��ɫ’s leadership as a research-intensive university that from its inception has understood the importance of an interdisciplinary approach in tackling complex, global problems. Connected Minds is particularly timely as we consider the implications of AI for creating a more equitable and inclusive world.” – Rhonda Lenton, 91��ɫ president and vice-chancellor

“91��ɫ is an international leader in interdisciplinary research involving artificial intelligence and other disruptive technologies, social justice, and human science like neuroscience, philosophy, and psychology. The government’s substantial investment will unite 91��ɫ’s incredible strengths with Queen’s health specialties to chart new territory in socially responsible, community-engaged research for a rapidly changing digital world.” – Amir Asif, 91��ɫ vice-president, Research and Innovation

“The current technological revolution will have transformative positive impacts, and likely unintended negative impacts, on humanity for generations to come. To predict these impacts and steer toward positive outcomes, one requires transdisciplinary expertise, multisector community engagement, and research and training at levels that can only occur in a large-scale program. We thank CFREF for providing Connected Minds with the resources to lead Canada and the world in this timely and critical enterprise.” – Doug Crawford, 91��ɫ Distinguished Research Professor in Neuroscience and inaugural Scientific Director of Connected Minds 

“New technologies are developing exponentially and systems like the law are simply not keeping up. 91��ɫ’s motto, tentanda via, the way must be tried, guides us, in everything we will do.  We believe our inclusive, interdisciplinary approach that aligns with the UN sustainable development goals makes 91��ɫ the perfect place for anticipating the way humans and machines will, and should, connect in an equitable society.  This way must be tried.” – Pina D’Agostino, director and founder of 91��ɫ’s IP Innovation Clinic at Osgoode Hall Law School and Vice Director of Connected Minds. 

“Connected Minds is informed by Indigenous perspectives and priorities to achieve outcomes that are culturally relevant and responsive to Indigenous ways of being and doing that impact how we think about and engage in life, health, and education. Our work will seek to address the unexpected consequences of technological innovation, like the growing digital divide for Indigenous communities to access remote health care, and issues of data sovereignty, ownership and digital colonialism.” – Sean Hillier, director of 91��ɫ’s Centre for Indigenous Knowledges & Languages and Associate Director of Connected Minds

“The Connected Minds project builds on a history of partnership and collaboration between Queen’s and 91��ɫ. Each institution brings unique but complementary research strengths to bear on the important challenges and opportunities that come with disruptive technologies and their impact on Canadian and global citizens.”— Nancy Ross, vice principal research at Queen’s University

“I look forward to working with Indigenous, community and industrial partners to develop more equitable and socially responsible research outputs for the benefit of all. I am also excited about the many educational and outreach opportunities that Connected Minds will produce – from school programs to graduate training and professional skills development. We want to democratize education and access to knowledge, with the aim of spreading a new culture of innovation for a more equitable, inclusive, and healthy society.” – Gunnar Blohm, Professor in Computational Neuroscience and Vice Director of Connected Minds

-30-

For more information:

91��ɫ U media contact:

Emina Gamulin

egamulin@yorku.ca

437-217-6362

Queen’s media contact:

Julie Brown

brown.julie@queensu.ca

343-363-2763

The post 91��ɫ leads $318.4M first-of-kind inclusive next-gen technology research initiative appeared first on News@91��ɫ.