Prof. Hana Okoro

Biography

Hana Okoro joined Veyra Institute in 2014 and has held the rank of full Professor since 2020. She received her PhD in computational neuroscience from the Alderton Graduate Institute of Neural Systems in 2009, where she studied the geometry of sensory representation in early visual cortex using dimensionality-reduction methods applied to population recordings. Following her doctorate she was a research fellow at the Helmar Center for Theoretical Biology for four years, extending her work to auditory and somatosensory systems and beginning a long collaboration with neural-engineering researchers. She studied biophysics and mathematics as an undergraduate at Vestmore University.

Okoro's laboratory develops and applies mathematical frameworks for understanding how neural circuits compute. Current emphases include: normative models that derive circuit motifs from computational objectives (predictive coding, efficient coding, energy constraints); data-driven latent-variable models of large-scale population activity; and learning rules — both biologically plausible and gradient-derived — that can explain the acquisition and consolidation of skills. A growing strand of the group's work bridges to neuromorphic engineering in collaboration with the Neural Engineering Group: translating computational principles derived from biology into hardware architectures. The group uses high-resolution spike-sorting data from the Institute's Advanced Microscopy Centre for model validation.

Okoro has published 55 peer-reviewed articles, has given keynote lectures at the Annual Meeting on Computational Neuroscience and the International Congress of Cognitive Systems, and currently serves as associate editor of the Journal of Theoretical Neurobiology. She directs the Veyra CNS Colloquium series and teaches the graduate course Statistical Models of Neural Circuits every autumn.

Research interests

Selected publications

1. Okoro H, Adeyemi S, Wirth M. "Low-dimensional manifolds of motor cortical population activity across learned tasks." Nature Neuroscience and Computation, 7(4): 310–327, 2024. VEYRA-DOI: 10.veyra/VX-2403
2. Okoro H, Steiner R. "Predictive coding in multi-area visual hierarchies: a normative derivation of lateral and feedback connections." PLOS Computational Biology, 19(8): e1011450, 2023. VEYRA-DOI: 10.veyra/VX-2309
3. Ikenna C, Okoro H, Halmstad L. "A biologically plausible learning rule for spiking neural networks derived from dendritic prediction errors." Journal of Theoretical Neurobiology, 44: 100–118, 2022. VEYRA-DOI: 10.veyra/VX-2217
4. Okoro H, Ferreira P. "Structured variability and task-relevant geometry in auditory cortex populations." Neuron Models and Circuits, 33(2): 88–103, 2021. VEYRA-DOI: 10.veyra/VX-2116
5. Okoro H. "Energy-efficient coding in recurrent circuits: a constrained optimisation framework." Neural Computation, 31(6): 1120–1148, 2019. VEYRA-DOI: 10.veyra/VX-1910
6. Wirth M, Okoro H, Ndalama K. "GPFA extensions for non-Gaussian observation models in large-scale spike-sorting data." Journal of Neuroscience Methods, 298: 45–57, 2018. VEYRA-DOI: 10.veyra/VX-1808

Current group members

Postdoctoral researchers

• Dr. Seun Adeyemi — low-dimensional dynamics in motor and frontal circuits
• Dr. Marta Wirth — latent-variable models for large-scale neural recordings
• Dr. Pedro Ferreira — multi-sensory integration and auditory-visual representations

Doctoral students

• Chidi Ikenna — biologically plausible learning in spiking networks (Year 3)
• Fumiko Tanabe — efficient coding and adaptation in the auditory pathway (Year 2)
• Amara Ndalama — geometry of task representations across cortical areas (Year 1)