Jakob H. Macke

The paper uses data-constrained RNNs to analyze neural activity in macaques and finds attractor dynamics underlie sequence working memory.

This paper introduces DECODE-Plex, a deep learning framework for multi-channel single-molecule localization microscopy, improving imaging speed and quality in dense, overlapping data.

The paper introduces JO_SCPLOWAXLEYC_SCPLOW, a framework for differentiable simulation of biophysical neuron models, enabling efficient parameter optimization and task-driven training of large-scale models.

This paper uses simulation-based inference to tune a finite-element model of the middle ear, providing probability distributions of parameters that reproduce experimental data and enabling objective diagnosis of hearing loss.

velotest assesses RNA velocity embeddings' faithfulness to high-dimensional data, revealing quality differences across methods and datasets.

PRIME predicts cortical excitability from EEG in real-time, enabling personalized TMS interventions by adapting to individual brain states.

The paper introduces a standardized ion channel model based on Hodgkin-Huxley formulation, demonstrating its ability to fit most voltage-gated models and enable parameter comparison and inference.

The paper develops a hybrid modeling approach to link ion channel gene expression to the electrophysiological properties of cortical neuron types using statistical and mechanistic models.

This study uses automated search algorithms to discover diverse synaptic plasticity rules that support memory tasks and network stability.

This paper introduces an SBI framework for diffusion MRI analysis, showing it provides fast and accurate parameter estimation and uncertainty mapping compared to classical Bayesian methods.

The paper examines whether the phase of neuronal firing reflects temporal order in sequence memory in humans and RNNs, finding that phase order does not match item order.

The paper introduces a VAE-based method to model conditional distributions between neural and behavioral data, validated on fly and monkey datasets.

The paper develops a biophysical model of the mouse outer retina to study contrast invariant coding, finding that nonlinearities in the retina enable robust classification despite contrast variations.

The paper introduces a method to predict synaptic connectivity from neural tissue reconstructions, showing three assumptions can explain connectivity in mouse and human datasets.

This study develops a machine learning framework to predict motor excitability states using EEG features, achieving 72% mean accuracy in healthy participants.

The paper investigates how mice distinguish self-motion from object motion using virtual reality and neural recordings from the superior colliculus, finding that different layers of the superior colliculus encode these signals differently.

This paper uses denoising diffusion probabilistic models to generate realistic neurophysiological time series, capturing both general and dataset-specific features across multiple species and recording modalities.

The paper introduces AutoMIND, a method for inverse modeling of neural circuits, revealing dynamic-dependent invariances in parameter configurations across different neural dynamics.

The paper decodes movie features like characters, settings, and visual transitions from neuronal population activity in the human medial temporal lobe, showing region-specific information encoding.

This paper shows that trained recurrent neural networks develop phase-locked limit cycles to support working memory through coupling with an external reference oscillation.

The paper shows that multiple distinct plasticity rules can produce similar network dynamics, highlighting the need to consider degenerate mechanisms in plasticity studies.

This paper introduces simulation-based Bayesian inference to efficiently identify generative models in connectomics by inferring wiring rule parameters from empirical data.

The paper investigates how metabolic cost influences neural circuit parameters in the crab pyloric network, finding energy-efficient circuits with diverse conductance combinations.

This paper introduces MNLE, a simulation-based inference method for decision-making models, showing it is more efficient and accurate than existing methods.

The paper shows that neuron morphology influences cortical network architecture by shaping recurrent or feedforward connectivity patterns based on dendritic and axonal properties.

This paper develops a videography-based method to estimate skeletal pose and joint kinetics in freely-moving rodents, validated by direct limb placement measurements.

The paper shows that dentate gyrus activity during immobility generates structured population patterns that are necessary for forming spatial memories.

This paper introduces a machine learning method using deep neural density estimators to perform Bayesian inference on neural dynamics models, demonstrating its application to various neural systems.

The paper benchmarks spike rate inference algorithms for two-photon calcium imaging data and finds that top-performing algorithms show high correlation despite diverse approaches.