Institution profile

University of Bern

Academic institutioneurope · ch
Official website
Research library236linked papers
Opportunities0open roles
Selected work

Representative Papers

Backpropagation through space, time, and the brain

Mar 25, 2024arXiv.org

How to achieve efficient credit assignment under spatiotemporal locality constraints in physical neural networks remains a fundamental challenge in neuromorphic computing. This paper introduces the Generalized Latent Equilibrium (GLE) framework, which for the first time couples energy minimization with neuron-level local mismatch dynamics to derive biologically plausible forward and backward continuous-time dynamics. By incorporating dendritic morphology modeling and membrane potential phase modulation, GLE implements spatiotemporal convolution and temporal reversal of feedback signals. Crucially, GLE relies exclusively on local synaptic plasticity—requiring no global timing coordination or external error broadcasting. Experiments demonstrate that, under strict locality constraints, GLE approximates the performance of backpropagation through time (BPTT), enables real-time online learning, incurs minimal memory overhead, and provides an interpretable, biologically realistic credit assignment mechanism for deep cortical networks.

8 citationsRead paper

A Practical Framework of Key Performance Indicators for Multi-Robot Lunar and Planetary Field Tests

Jan 28, 2026

This study addresses the lack of a unified, science-driven performance evaluation framework for multi-robot planetary exploration, which hinders meaningful cross-system comparisons. To bridge this gap, the work proposes the first science-oriented key performance indicator (KPI) framework tailored to three realistic lunar multi-robot cooperative scenarios. The framework is hierarchically structured around three dimensions—efficiency, robustness, and accuracy—and has been deployed and validated in field trials. It effectively narrows the divide between engineering metrics and scientific objectives: efficiency and robustness metrics prove readily applicable, while accuracy metrics remain constrained by the difficulty of obtaining ground-truth data. Overall, the framework serves as a standardized tool to advance the evaluation and optimization of robotic systems for planetary exploration.

1 citationsRead paper

Multihead self-attention in cortico-thalamic circuits

Apr 08, 2025

Bridging biological plausibility with Transformer computation remains a fundamental challenge in computational neuroscience and AI. Method: We establish a mechanistic, differentiable mapping between cortical-thalamic circuits and multi-head self-attention (MHSA) via computational neuroscience modeling, circuit-level equivalence analysis, and analytical derivation of gradients for linear MHSA. Contribution/Results: We present the first mathematically equivalent cortical-thalamic circuit model of MHSA; propose a functional division hypothesis wherein superficial and deep pyramidal neurons within cortical microcolumns encode attention masks and modulated values, respectively, endowed with differentiable learning; and validate the model across scales using electrophysiological and anatomical data—demonstrating high structural–functional correspondence and yielding analytically tractable, learnable gradients under token-wise MSE loss. Our work uniquely bridges biologically realistic neural circuits with Transformer mechanisms through a testable, mechanistic, and differentiable framework.

1 citationsRead paper

Weight transport through spike timing for robust local gradients

Mar 04, 2025

Backpropagation in deep spiking neural networks (SNNs) relies on symmetric weight connections, conflicting with biological locality constraints and hardware implementation requirements. Method: We propose spike-based alignment learning (SAL), a biologically plausible training mechanism grounded in spike-timing statistics. SAL integrates STDP, dual-mode Hebbian/anti-Hebbian plasticity, and intrinsic neuronal noise to adaptively align asymmetric feedforward–feedback weights—without requiring weight symmetry or explicit backward weight transmission—thereby recovering accurate local gradients. The model employs probabilistic spiking neurons and a hierarchical architecture inspired by cortical microcircuits. Contribution/Results: SAL significantly improves convergence accuracy toward target distributions and enables automatic alignment of feedback weights across multiple layers. Local error estimates achieve accuracy comparable to ideal backpropagation. The method satisfies key neurobiological constraints while demonstrating robustness under realistic computational conditions.

1 citationsRead paper

Nonparanormal Modeling Framework for Prognostic Biomarker Assessment with Application to Amyotrophic Lateral Sclerosis

Feb 28, 2025

In amyotrophic lateral sclerosis (ALS), the prognostic performance of biomarkers—such as serum neurofilament light chain (NfL)—varies dynamically over time and across patient subgroups (e.g., age, site of onset), yet existing time-dependent ROC methodologies often neglect covariate adjustment. Method: We propose the first joint modeling framework based on the nonparanormal transformation, simultaneously characterizing censored survival times and non-normally distributed biomarker trajectories. By employing a copula to model dependence structures and explicitly adjusting for confounding covariates, our approach enables covariate-specific, time-dependent ROC estimation. Contribution/Results: Our method relaxes restrictive normality assumptions, flexibly capturing biomarker–covariate interactions and supporting individualized, dynamic prognostic assessment. Empirical analysis demonstrates that NfL’s predictive utility is significantly modulated by both temporal windows and clinical subgroups. When applied to clinical trial design, it improves patient stratification accuracy and reduces required sample sizes by 20%–35%.

1 citationsRead paper
Recent publications

Latest Papers