Gave a seminar on my research on local learning rules for DNNs at the University of Twente!

I recently had the privilege of presenting my research at the Data Management and Biometrics Group at the University of Twente. This seminar was an exciting opportunity to showcase my work on local learning rules for deep neural networks, particularly two novel approaches designed to train artificial spiking and non-spiking neural networks with minimal computational complexity and memory requirements.

In the session, I shared the latest results on the two local learning rules I developed, S-TLLR and LLS, for spiking and non-spiking networks. I highlighted their biological inspirations, practical applications, and potential for enabling efficient, scalable on-device learning. We had a lively discussion about the challenges and future directions in on-device learning, and I thoroughly enjoyed connecting with such dedicated professionals.

A special thank you to Prof. Estefanía Talavera for her invaluable mentorship over the past two years and for inviting me to speak. Meeting in person after years of virtual mentorship made this experience especially meaningful.

Thank you, University of Twente, for the warm welcome. I’m grateful for the connections and conversations sparked by this seminar.