As sistant Professor of Neuroscience
WashU Medicine
Gaia Tavoni\, PhD
As
sistant Professor of Neuroscience
WashU Medicine
Large-scale\, high-resolution neural recordings allow us to ana lyze neural activity simultaneously across sensory\, decision-making and mo tor systems. Yet these systems are typically studied in isolation\, leaving open the central question of how distributed brain networks jointly encode \, integrate and predict information to guide behavior. At the same time\, theoretical accounts of neural coding often assume distinct computational o bjectives — maximizing information transmission (efficient coding)\, minimi zing prediction errors (predictive coding) or optimizing performance for sp ecific tasks (task-driven coding). How these frameworks relate to one anoth er and whether a single unifying principle can explain neural computation a cross brain areas and behavioral contexts remains unclear.
One line of research in my lab aims to establish a unified mathematical theory of neural coding that addresses these gaps. In this ta lk\, I will present a framework we recently developed that generalizes effi cient coding to multimodal networks — circuits that integrate independent s treams of sensory and motor information. This theory: (a) provides a unifie d account of diverse multisensory and sensorimotor phenomena\, including mo tor influences on auditory processing and cross-modal interactions among vi sual\, auditory\, somatosensory and olfactory systems\; (b) shows that pred ictive computations arise naturally from efficient codes in low-noise regim es\, linking two major coding principles within a single mathematical frame work\; (c) uncovers a circuit-level algorithm for predictive coding that di ffers fundamentally from previous proposals (for which empirical evidence h as so far been limited)\, and situates predictive coding within a broader i nterpretative landscape\; and (d) recovers unimodal coding as a limiting ca se\, connecting classical results to a more general multimodal formulation. Finally\, I will show how this theory maps onto biologically plausible lea rning dynamics in sparse neural networks.
Together\, these results establish a foundation that we will extend to addi tional circuit motifs\, noise regimes and behavioral contexts in pursuit of a general theory of neural coding that bridges normative objectives\, algo rithmic mechanisms and circuit-level implementations across the brain.
CATEGORIES:Seminar Series LOCATION:Neuroscience Research Building Auditorium GEO:38.635602;-90.254892 ORGANIZER;CN="Shea":MAILTO:shea.stewart@wustl.edu URL;VALUE=URI:https://neuroscience.wustl.edu/events/event/department-of-neu roscience-seminar-gaia-tavoni/ END:VEVENT BEGIN:VTIMEZONE TZID:America/Chicago BEGIN:STANDARD TZOFFSETFROM:-0500 TZOFFSETTO:-0600 DTSTART:20251102T070000 TZNAME:CST END:STANDARD END:VTIMEZONE END:VCALENDAR