ARC@ORU: On the Structure of Neuromorphic Computation: Spikes, Memory and Reset
17 juni 2026 13:15 – 14:00 Visual Lab, ARC
ARC inbjuder till ett forskningsseminarium med Ayca Özcelikkale, docent och universitetslektor vid Institutionen för elektroteknik; Signaler och system, Uppsala universitet. Seminariet sker inom ramen för en workshop med "the Unconventional Computing Nordic Research Network" (UniCoRN).
ARC@ORU Research Seminar Series
On the Structure of Neuromorphic Computation: Spikes, Memory, and Reset
This is a Key Note seminar within the kick-off workshop for the Unconventional Computing Nordic Research Network, UniCoRN.
About the talk
Neuromorphic computing is an emerging paradigm for low-latency, energy-efficient information processing, in which brain-inspired models such as spiking neural networks play a central role. At the same time, deep state space models have recently shown strong performance in temporal modelling. Although these approaches come from different research traditions, both implement stateful transformations of temporal input sequences. This talk presents a principled correspondence between these models through a dynamical systems approach. It also discusses broader questions about reset, stability and memory in learning, and what these may mean for the design of neuromorphic systems.
Speaker
Ayça Özçelikkale is Associate Professor at the Department of Electrical Engineering at Uppsala University, Sweden. Before joining Uppsala University, she held a postdoctoral position at Chalmers University of Technology and visiting positions at Ericsson Research, Queen’s University and EURECOM. Her research includes signal processing, machine learning, neuromorphic computing and communications.