March 16, 2026 • Cascais, Portugal

Edge of Tomorrow: Theory and Practice of Closed-Loop Neural Control

Moving from passive recording to active control: bridging the gap between control theory algorithms and experimental implementation.

About the Workshop

The workshop aims to catalyze the next wave of systems neuroscience research by moving from passive recording of neural activity to active closed-loop control. This workshop will bridge the gap between control theorists developing algorithms and experimentalists navigating the challenges of implementing these algorithms in practice.

While past Cosyne workshops have explored the constituent elements of dynamical system identification, active inference, and optimal control of neural population dynamics, our proposed workshop offers a distinct and timely synthesis focused on the translational potential of the emerging closed-loop techniques.

We will specifically emphasize methods applicable to human electrophysiology, where the goal is to actively probe and control neural dynamics with precision, maximizing information gain within limited experimental sessions. This workshop will create a novel forum that bridges the gap between control theorists developing algorithms and experimentalists facing the practical hurdles of implementing these systems, thereby embodying the Cosyne ethos of a tight, functional integration of theory and experiment to ask fundamentally new questions about neural computation.

Invited Speakers

Leading experts at the intersection of theory and experiment

Simon Jacob

Simon Jacob

Technical University of Munich (TUM)

Human Cortex Microstimulation

"Single-neuron and neuronal circuit dynamics of intracortical electrical microstimulation in human association cortex"

Luke Bashford

Luke Bashford

Newcastle University

BCI Frontoparietal Network

"Closed loop recording and intracortical microstimulation of the frontoparietal network in humans with an implanted brain-computer interface"

Georgia Koppe

Georgia Koppe

Heidelberg University / ZI Mannheim

Dynamical Systems Data-driven Models

"Tractable data-driven models for dynamical system reconstruction and control of brain and behavior"

Mikito Ogino

Mikito Ogino

The University of Tokyo

Optimal Design Controllability

"Optimal design of stimulation for neural system identification and controllability estimation"

Kyle Johnsen

Kyle Johnsen

Georgia Tech

Optogenetics Simulations

"How many channels do we need to control latent neural dynamics in real time? Insights from a high-throughput survey"

Memming Il Park Hyungju Jeon

Memming Il Park & Hyungju Jeon

Champalimaud Research

Real-time Learning Neural Dynamics

"Planning to Learn Neural Dynamics in Real Time"

Schedule

Monday, March 16, 2026

Morning Session (9:30 - 12:20)

09:30 - 10:00

Opening Remarks

Nina Kudryashova, Melina Müller, Matthias Hennig & Luke Bashford

10:00 - 10:40

How many channels do we need to control latent neural dynamics in real time?

Kyle Johnsen

Coffee Break (10:40 - 11:00)
11:00 - 11:40

Single-neuron and neuronal circuit dynamics of intracortical electrical microstimulation in human association cortex

Simon Jacob

11:40 - 12:20

Tractable data-driven models for dynamical system reconstruction and control of brain and behavior

Georgia Koppe

Lunch Break

12:20 - 15:00

Afternoon Session (15:00 - 18:00)

15:00 - 15:40

Closed loop recording and intracortical microstimulation of the frontoparietal network in humans

Luke Bashford

15:40 - 16:20

Optimal design of stimulation for neural system identification and controllability estimation

Mikito Ogino

Coffee Break (16:20 - 16:40)
16:40 - 17:20

Planning to Learn Neural Dynamics in Real Time

Memming Park / Hyungju Jeon

17:20 - 18:00

Panel Discussion

Moderated by Organizers with All Speakers

Organizers

Nina Kudryashova

Nina Kudryashova

University of Edinburgh

Contact
Melina Müller

Melina Müller

University of Edinburgh

Contact
Matthias Hennig

Matthias Hennig

University of Edinburgh

Contact
Luke Bashford

Luke Bashford

Newcastle University

Contact