Research

I build computational and mathematical models of how neural circuits assemble and reorganise. Currently I'm focussing on early cortical development.

Schematic of the excitatory and inhibitory microcircuit of mouse
somatosensory cortex at four stages of postnatal development. — The cortex is spontaneously active long before sensory experience arrives and begins to shape it, and that spontaneous activity moves through a sequence of distinct regimes (from large synchronous bursts, through to the sparse, decorrelated firing of the adult) while the circuitry assembles. I model how that assembly schedules these activity transitions, and how the cortex arrives at the regime it processes the world in.

Other systems I've worked on include:

Representational drift in mouse visual cortexMark Hübener, MPIBI, Munich
Graph rules for chaos in spiking networksMBL, Woods Hole
Sparsity in low-rank recurrent networksSrdjan Ostojic, ENS, Paris
Beta oscillations in macaque premotor cortexVincent Hakim, ENS, Paris
Efficient coding in the retinaMatthew Chalk, Institut de la Vision, Paris
Collective behaviour in fruit fliesGonzalo de Polavieja, Champalimaud, Lisbon
Unconscious bias in human decision-makingKatharina Henke, Bern

PhD researcher since 2022 in the lab of Julijana Gjorgjieva, Technical University of Munich. I did my MSc at ENS Paris, and BSc at the University of Bristol. Between them, I took a year off to travel around Europe to meet with researchers, artists and thinkers, and piece together a map of computational neuroscience "as seen from the inside". I publish as Elizabeth Herbert.

Publications

2025Lakhera, S.*, Herbert, E.*, & Gjorgjieva, J. Modeling the emergence of circuit organization and function during development. Cold Spring Harbor Perspectives in Biology 17(2), a041511. *co-first authors
2024Bauer, J., Lewin, U., Herbert, E., Gjorgjieva, J., Schoonover, C., Fink, A., Rose, T., Bonhoeffer, T., & Hübener, M. Sensory experience steers representational drift in mouse visual cortex. Nature Communications 15, 9153.
2022Herbert, E., & Ostojic, S. The impact of sparsity in low-rank recurrent neural networks. PLOS Computational Biology 18(8), e1010426.
2017Ruch, S., Herbert, E., & Henke, K. Subliminally and supraliminally acquired long-term memories jointly bias delayed decisions. Frontiers in Psychology 8, 1542.