cv
What I've done with my life so far.
Basics
| Name | Gabriel Béna |
| Label | Postdoctoral Researcher in Neuromorphic Computing |
| bena@ini.uzh.ch | |
| Summary | Postdoctoral researcher at the Institute of Neuroinformatics (INI), Zurich, working at the intersection of computational neuroscience, machine learning, and neuromorphic engineering. Interested in how complex, adaptive computations emerge from simple local parts — through modularity, self-organisation, and the structure–function relationship in neural networks. |
Work
-
2026.05 - Present Postdoctoral Researcher
Institute of Neuroinformatics (INI), UZH / ETH Zurich
Postdoc in Melika Payvand's lab, exploring how spatial and energy constraints in memristive neuromorphic hardware shape the emergence of modular, compositional computation.
- Spatially-embedded recurrent networks on memristor-based neuromorphic substrates
- Energy- and locality-driven emergence of structural and functional modularity
- Compositional learning and generalisation under physical hardware constraints
-
2023.10 - 2025.11 Neuromorphic Researcher & Developer
SpiNNcloud Systems
Part-time researcher based in Dresden, Germany, working on neuromorphic computing applications and SNN software for the SpiNNaker2 platform.
- Designed and deployed deep SNNs on the SpiNNaker2 neuromorphic chip
- Built hybrid pipelines combining conventional and neuromorphic compute
- Tackled optimisation and physics problems as neuromorphic applications
- Active developer of the py-spinnaker2 software stack
-
2021.01 - 2026.03 PhD Candidate
Imperial College London
PhD in the Neural Reckoning Group under the supervision of Dan Goodman, on the structure–function relationship in neural networks and on self-organisation as a route to robust, distributed computation. Full thesis: Physical Constraints and Functional Demands shape Modular Neuromorphic Intelligence (doi:10.25560/128990).
- Showed that structural modularity alone does not guarantee functional specialisation: the structure–function link emerges only under specific resource constraints and input statistics (Nature Communications, 2024)
- Trained continuous Neural Cellular Automata as a universal computational medium, demonstrating learnable matrix-algebra primitives and MNIST-classifier emulation directly within the cellular state
- Introduced Self-Organising Digital Circuits: a Topology-Masked Transformer that grows and self-repairs Boolean logic circuits via purely local message passing, with no global backpropagation at deployment
- Co-developed the SNN software stack for SpiNNaker2 and contributed to multimodal-integration studies in Spiking Neural Networks
-
2020.06 - 2020.12 ML / Neuromorphic Intern
Prophesee
Internship in Paris, France working on event-based cameras and neural networks.
- Trained deep spiking neural networks for event-based cameras
- Created spiking auto-encoders for denoising and compressing tasks
- Worked on end-to-end event-based systems using Intel's LOIHI processor and Prophesee's event-based cameras
Education
-
2019.01 - 2020.01 Toulouse, France
Master
Université Paul Sabathier
Operational Research
- Applied Mathematics
- Informatics
- Engineering
- Optimization
- Modelization
- Algorithms
-
2016.01 - 2020.01 Toulouse, France
Master of Science
ISAE-Supaero
Data and Decision Science, Machine Learning applied to Space Systems
- Data Science
- Decision Science
- Machine Learning
- Space Systems Operation
- Space Systems Conception
-
2014.01 - 2016.01 Paris, France
Skills
| Programming | |
| Python | |
| C | |
| Java |
| Deep Learning | |
| ANNs | |
| SNNs | |
| PyTorch | |
| JAX |
| Machine Learning | |
| Supervised Learning | |
| Unsupervised Learning | |
| Reinforcement Learning | |
| Statistical Modeling |
Languages
| French | |
| Native |
| English | |
| Fluent |
| Spanish | |
| Intermediate |
Interests
| Sports | |
| Skiing | |
| Surf | |
| Rugby |
| Music | |
| Saxophone | |
| Guitar |
| Travel | |
| Morocco (5 years) | |
| England (4 years) | |
| Japan (6 months) | |
| Colombia (6 months) |
| Solarpunk |