Seminars

Functional connectomics reveals general wiring rule in mouse visual cortex

Charles Bordenave revisits von Neumann and Goldstine's proposal to model numerical computing errors as probabilistic noise. This perspective helps explain large errors in computed eigenvalues of badly conditioned matrices. Toeplitz matrices provide a case study: they occur throughout mathematics, physics, and engineering, are highly sensitive to small perturbations, and possess a rich spectral theory that supports precise descriptions under random perturbations. The talk is based on joint work with Mireille Capitaine and François Chapon.

Connectomic traces of Hebbian plasticity in the entorhinalhippocampal system

Different brain cell types exhibit distinct metabolic signatures that link energy economy to cellular function. Astrocytes and neurons, for instance, diverge dramatically in their reliance on glycolysis versus oxidative phosphorylation, underscoring that metabolic fuel efficiency is not uniform across cell types. A key factor shaping this divergence is the structural organization of the mitochondrial respiratory chain into supercomplexes. Specifically, complexes I (CI) and III (CIII) form a CI–CIII supercomplex, but the degree of this assembly varies by cell type. In neurons, CI is predominantly integrated into supercomplexes, resulting in highly efficient mitochondrial respiration and minimal reactive oxygen species (ROS) generation. Conversely, in astrocytes, a larger fraction of CI remains unassembled, freely existing apart from CIII, leading to reduced respiratory efficiency and elevated mitochondrial ROS production. Despite this apparent inefficiency, astrocytes boast a highly adaptable metabolism capable of responding to diverse stressors. Their looser CI–CIII organization allows for flexible ROS signaling, which activates antioxidant programs via transcription factors like Nrf2. This modular architecture enables astrocytes not only to balance energy production but also to support neuronal health and influence complex organismal behaviors.

Cellular crosstalk is an essential process during brain development and is influenced by numerous factors, including cell morphology, adhesion, the local extracellular matrix and secreted vesicles. Inspired by mutations associated with neurodevelopmental disorders, we focus on understanding the role of extracellular mechanisms essential for the proper development of the human brain. Therefore, we combine 2D and 3D in vitro human models to better understand the molecular and cellular mechanisms involved in progenitor proliferation and fate, migration and maturation of excitatory and inhibitory neurons during human brain development and tackle the causes of neurodevelopmental disorders.

Partha Dasgupta considers how economic progress should be measured during an ecological crisis. Drawing on his work on natural capital, he examines an approach to economics that assigns value to nature alongside other assets. The discussion asks how changing the way economies value the natural world could change the way they protect it. Dasgupta connects economic measurement with environmental sustainability and the challenge of recognising nature’s contribution to human well-being.

Jeroen Zuiddam examines whether repeated tasks permit economies of scale, focusing on fast matrix multiplication and graph Shannon capacity as unresolved direct-sum problems. The talk introduces Strassen's asymptotic-spectrum duality, originally developed for asymptotic tensor rank and later applied to Shannon capacity and other problems. It then explores approximation by sequences of easier instances: suitable notions of convergence, construction of the sequences, asymptotic-spectrum distance, and approaches using polynomials and algebraic geometry. Based on joint work with David de Boer, Pjotr Buys, Sven Polak, Matthias Christandl, Koen Hoeberechts, Harold Nieuwboer, and Péter Vrana.

Research articles and laboratory protocol organization often lack detailed instructions for replicating experiments. protocols.io is an open-access platform where researchers collaboratively create dynamic, interactive, step-by-step protocols that can be executed on mobile devices or the web. Researchers can easily and efficiently share protocols with colleagues, collaborators, the scientific community, or make them public. Real-time communication and interaction keep protocols up to date. Public protocols receive a DOI and enable open communication with authors and researchers to foster efficient experimentation and reproducibility.

Distinct synaptic plasticity rules operate across dendritic compartments in vivo during learning

Neuroplasticity is essential for the establishment and strengthening of neural circuits. Repetitive transcranial magnetic stimulation (rTMS) is commonly used to modulate cortical excitability and shows promise in the treatment of some neurological disorders. Low intensity magnetic stimulation (LI-rTMS), which does not directly elicit action potentials in the stimulated neurons, have also shown some therapeutic effects, and it is important to determine the biological mechanisms underlying the effects of these low intensity magnetic fields, such as would occur in the regions surrounding the central high-intensity focus of rTMS. Our team has used a focal low-intensity (10mT) magnetic stimulation approach to address some of these questions and to identify cellular mechanisms. I will present several studies from our laboratory, addressing (1) effects of LIrTMS on neuronal activity and excitability ; and (2) neuronal morphology and post-lesion repair. The ensemble of our results indicate that the effects of LI-rTMS depend upon the stimulation pattern, the age of the animal, and the presence of cellular magnetoreceptors.

Matrix multiplication underpins scientific computing and artificial intelligence, yet practical numerical libraries and hardware accelerators commonly retain the classical cubic-time algorithm despite decades of subcubic theoretical advances. This talk reviews the effort to make faster multiplication algorithms useful in practice. It examines why arithmetic complexity alone does not determine performance: some algorithms require enormous matrices or incur large hidden constants, while communication costs, numerical stability and the match between software and hardware create additional obstacles. The historical perspective connects asymptotic algorithm design to actual performance and power consumption.