Attention & Perception Talk Series

Links to the speaker's website and and talk information will be added as the semester goes along. Talks are online (Zoom link) unless otherwise specified. In-person talks are held in Room 819. As scheduling is ongoing, some of the "free" slots may be filled. All talks are on Tuesdays at 2 PM CT.

Fall 2026

  • 08/25 — First meeting (area outing; no speaker)
  • 09/01 — No speaker scheduled
  • Aalap Shah (Yale) — A computational framework for understanding embodied plans

    September 8th, 2026 (online)

    Reasoning about people's plans and goals based on their actions is fundamental to human social interaction and essential for cooperative, embodied artificial intelligence. Formalizing this capacity in naturalistic, embodied settings has been a longstanding challenge. So, in this talk, I will present GenEA, a novel agent representation capable of reasoning over multi-step, long-horizon embodied plans. At the heart of GenEA lies a multilevel state-space that couples the coarse biomechanics of an articulated body and its interactions with nearby objects by extending the task-and-motion planning framework from robotics with simulation-based intuitive physics. By embedding this state-space in a generative model, GenEA yields a joint probability distribution over goals and embodied plans. Across several case studies spanning qualitatively distinct domains of object manipulation and navigation, I will show that GenEA, through simple transformations of this joint distribution, recapitulates detailed patterns of how humans reason about embodied plans. Notably, GenEA consistently outperforms strong alternative models, including plausible heuristics, standard task-and-motion planning, and a state-of-the-art vision-language model. GenEA thereby offers a new perspective on the architecture of human reasoning and social cognition, suggesting that embodied artificial intelligence systems seeking to seamlessly interact with humans should adopt GenEA-like generative models of action planning.

  • Anjana Lakshmi (Indiana University) — The Hyper-Typicality Effect: The typical face is not average

    September 15th, 2026 (in-person)

    In this work, we challenge prevailing accounts of social categorization that assume ethnic face typicality is anchored in central tendency — that the typical face for an ethnic group is the average face. Instead, we find that people represent group-typical faces as hyper-typical, with caricatured group-specific features — an exaggeration of the group's "essence". In three studies, we use computational modeling together with face-morphing techniques to compute a measure of physical group typicality that locates any face on a continuum from average to exaggerated on category-defining features. Using participants' typicality ratings of these faces, we identify where subjective typicality is maximal on this continuum. Across different computational approaches and research designs, we find that the typical face is not average: ethnic face typicality is systematically displaced towards caricatured exemplars with exaggerated group-differentiating features. We refer to this phenomenon as the Hyper-Typicality Effect (HTE).

  • Ed Awh (UChicago) — Working memory storage goes beyond the mere maintenance of information

    September 22nd, 2026 (in-person)

    Although past neural studies of working memory (WM) have focused on stimulus-specific activity that tracks the stored feature values, a separate line of evidence has revealed neural signals that track the number of items in WM, independent of the contents of those items. Thus, a common neural signature of WM load has been identified for highly distinct visual features, and even across visual and auditory sensory modalities. Our working hypothesis is that these content-independent load signals reflect the operation of spatiotemporal "pointers" that enable the binding of stored items to the surrounding event context, a process that is critical for precise access online memories despite high levels of proactive interference. Critically, there is growing evidence that this pointer operation is functionally distinct from the deployment of covert spatial attention, even though spatial attention often requires the maintenance of location information in the absence of external cues. In line with this dissociation, we show that WM activity reaches a plateau when set size exceeds WM capacity, while activity tracking spatial attention rises monotonically with no such limit. These findings motivate a refined definition of WM storage that acknowledges how item-based storage in WM goes beyond the mere maintenance of information.

  • Frank Durgin (Swarthmore) — Evidence that calibration of precursors of numerosity perception is responsible for large numerosity aftereffects

    September 29th, 2026 (online)

    In 1995, I reported a new aftereffect of adaptation to texture that seemed to change the perceived numerosity of dots in a texture patch. We argued that it did so by altering their perceived density. Importantly, the adaptation affected the perceived variance of texture fields, by seeming to reduce the perceived presence of high densities. We additionally showed that the effect could not be attributed to spatial frequency adaptation. In 2008 Burr and Ross argued that they had discovered a new number aftereffect, arguing that if number was adapted, then number was a directly sensed property of the world. In this talk I'll review recent work from my lab (3 pre-registered sets of studies) showing (1) that much of what has been claimed to be "number adaptation", can be traced to the adaptation of precursor dimensions that likely feed into numerosity perception like texture density and patch size, (2) that texture density perception is more precise than number perception, consistent with its hypothesized role as a precursor dimension, and (3) that, in particular, a study by Desimone et al. (2020) in Psych Science claiming to resolve the adaptation debate in favor of number adaptation went wrong because they failed to consider size adaptation. If there is time, I'll also describe a new study showing that it is retinal density, rather than perceived density that drives adaptation.

  • Sebastiaan Mathôt (U. of Groningen, Netherlands) — Beyond arousal: pupil fluctuations, neural activity, and behavior

    October 6th, 2026 (online)

    Spontaneous fluctuations in pupil size are correlated with neural activity and behavior in many ways. These relationships are often attributed to a single underlying cause: arousal. This unitary account disregards other potential underlying causes. To address this gap, Veera Ruuskanen and I propose a comprehensive framework of four mechanisms that jointly underlie these relationships. In addition to arousal, we consider optical effects, brightness constancy, and self-generated visual stimulation as factors driving the complex and context-dependent relationships between pupil size, neural activity, and behavior. Taken together, we argue that pupil fluctuations should be considered beyond arousal: as a functionally important, yet still poorly understood, component of visual processing.

    • Ruuskanen, V., & Mathôt, S. (2026). Beyond arousal: Pupil fluctuations, neural activity, and behavior. Trends in Cognitive Sciences. https://doi.org/10.1016/j.tics.2026.07.008
    • Mathôt, S., Dimigen, O., Karsilar, H., Ruuskanen, V., Weiden, D., & Vilotijević, A. (2026). Pupil constriction causes activity in the human retina and visual system. Psychophysiology. https://doi.org/10.1111/psyp.70362
  • 10/13 — Cyrus Yang; in-person
  • 10/20 — Hong Nguyen (New School); in-person
  • 10/27 — Federico Sorcini; in-person
  • 11/03 — Harini Sankar; in-person
  • 11/10 — Yifan Ding; in-person
  • 11/17 — Siddharth Munupatrula; in-person
  • 11/21-11/30 — Fall Break
  • 12/01 — Cyrus Yu; in-person
  • 12/08 — Cindy Xiong Bearfield (Georgia Tech); in-person