Intuitive physical reasoning in brains, minds, and machines

Organizers: RT Pramod¹, Nancy Kanwisher¹; ¹MIT

How do humans understand the physical scene in front of them, predict what will happen next, and plan actions? This “intuitive physics” ability arises early in infancy and engages a systematic set of parietal and frontal brain regions in adults. At the same time, physical reasoning represents an ongoing challenge in robotics, where the hardest problem is not movement itself, but understanding the physical world in order to plan movement. This symposium brings together cognitive scientists, neuroscientists, and computational modelers to present recent findings and discuss competing hypotheses of how humans understand and interact with the physical world. More…

Rhythms of vision: How neural oscillations structure perception and attention

Organizers: April Pilipenko¹, Jason Samaha¹; ¹University of California, Santa Cruz

The brain is dominated by oscillatory activity; This symposium highlights emerging evidence that visual perception and attention are governed by these rhythmic neural dynamics. Across six talks, speakers will demonstrate how oscillatory and aperiodic activity shape temporal integration, sensory encoding, visuospatial processing, and the resolution of neural competition. By spanning alpha-, beta-, and theta-band mechanisms and integrating methods using EEG, MEG, and neurostimulation, the symposium reveals how endogenous brain rhythms structure perceptual decisions and attentional states. Together, these perspectives offer a unified view on how oscillatory dynamics orchestrate perceptual processing and are linked to visual cognition. More…

What makes a task naturalistic? Towards a theory-inspired approach to studying visually-guided behaviour

Organizers: Jolande Fooken¹, Constantin A. Rothkopf¹; ¹Centre for Cognitive Science, Technical University Darmstadt, Germany

In natural visual behaviour, perception, memory, and action are inextricably linked to the goals of the task at hand. For example, when attending VSS, we actively search for visual information while reading abstracts, we use visual signals to guide our movements, and we rely on vision to read social cues. Our perception of the conference will also be shaped by expectations and memories. To deepen our understanding of natural visual processing, vision scientists should design tasks that capture the flexibility and diversity of visually-guided behaviour. In this symposium, we will highlight key features that make tasks naturalistic and behaviourally relevant. More…

Reimagining the binding problem(s) for the 21st century

Organizers: JohnMark Taylor¹, Seda Karakose-Akbiyik²; ¹Columbia University, ²Stanford University

Modern vision science faces a “methodological binding problem”: disparate subfields employing widely varying methods, paradigms, and terminology have enriched our understanding of different facets of the binding problem, but have not yielded integrated neurocognitive models of how the visual system combines features, or even consensus on the key outstanding questions. This symposium synthesizes novel behavioral, neural, and computational evidence that challenges us to reconceptualize core components of the binding problem—or problems—from the ground up, updating the broader VSS community on the state of the art and charting out future interdisciplinary directions on this fundamental challenge for biological and artificial vision. More…

Beyond prediction: Interpretable brain encoding models for understanding vision

Organizers: Zitong Lu¹, Apurva Ratan Murty²; ¹Massachusetts Institute of Technology, ²Georgia Institute of Technology

After a decade of leaderboard races and now near-ceiling performance on human fMRI data, many models “predict” visual brain data well. What, if anything, does this success reveal about how the visual system actually works? This symposium gathers early-career researchers pushing predictive models toward deeper and more precise understanding of human brains. Talks span multiple neural modalities (fMRI, EEG, single-neurons) and diverse encoding-model methods (explainable AI, generative modeling, circuit analysis to cognitive experiments). Together, these talks argue prediction is necessary but insufficient, and interpretable encoding models are our best bet for turning brain-like performance into genuine explanations of visual processing. More…

The quest for 'the average person' in vision science: Promise and pitfalls

Organizers: Jeremy Wilmer¹, Michael Herzog²; ¹Wellesley College, ²Swiss Institute of Technology (EPFL)

Human minds and brains are complex, and our measurements inevitably carry noise. As a result, vision science—and human research more broadly—typically relies on averaging across participants. Over time, this strategy yields ever-sharper estimates of the average observer. But is this the ultimate goal of vision science? This symposium takes a critical look at our reliance on the “average person” in vision research. When does averaging illuminate core mechanisms, and when does it blur or even distort understanding? Drawing on examples from diverse domains, our speakers will showcase both the power and the pitfalls of this longstanding analytic convention. More…