NBIO Presents: Asim Iqbal, PhD, (Tibbling Technologies)

NeuroAI for Generalised Neurological Discovery

The mammalian brain exhibits a remarkable ability to generalize to novel tasks, sensory modalities, and environments using limited data and high energy efficiency. In contrast, current AI systems rely on large-scale datasets, operate within fixed input modalities, and fail to build robust cross-domain representations, resulting in poor generalization and performance degradation when adapting to new tasks. In this seminar, I will introduce a NeuroAI architecture grounded in principles of neural computation, designed to identify and embed these mechanisms within multimodal transformer models for neuroscientific applications. This architecture supports few-shot generalization across a range of brain disorders, enabling applications in mechanistic discovery, personalized rehabilitation, and targeted therapeutic strategies