Computational Neuroscience
Computational models of the brain, neural coding, and brain-computer interfaces.
q-bio.NC · 115 papersFoundation models for discovering robust biomarkers of neurological disorders from dynamic functional connectivity
RE-CONFIRM evaluates biomarker robustness from brain foundation models, and Hub-LoRA improves their ability to identify neurobiologically faithful biomarkers.
Directional Confusions Reveal Divergent Inductive Biases Through Rate-Distortion Geometry in Human and Machine Vision
This paper uses directional confusions and Rate-Distortion geometry to reveal distinct inductive biases in human and machine vision.
Modulating Cross-Modal Convergence with Single-Stimulus, Intra-Modal Dispersion
A new method measures single-stimulus intra-modal dispersion, revealing it significantly modulates cross-modal convergence between vision and language models.
Hierarchical organization of critical brain dynamics
A study reveals how brain criticality signatures are hierarchically organized and task-modulated, linking collective neural dynamics to brain architecture.
Only Brains Align with Brains: Cross-Region Alignment Patterns Expose Limits of Normative Models
This paper introduces 'alignment patterns' to improve brain-model alignment benchmarks, revealing current methods' limitations and a need for stronger evidence.
Micro-DualNet: Dual-Path Spatio-Temporal Network for Micro-Action Recognition
Micro-DualNet is a dual-path spatio-temporal network that improves micro-action recognition by adaptively processing diverse spatial and temporal characteristics.
Response time of lateral predictive coding and benefits of modular structures
This paper reduces response time in Lateral Predictive Coding (LPC) systems and shows modular structures are equally effective with fewer connections.
Modelling time-order effects in haptic perception with a Bayesian dynamical framework
This paper introduces a Bayesian dynamical model to explain time-order effects in haptic perception, reproducing biases and individual variability.
OmniMouse: Scaling properties of multi-modal, multi-task Brain Models on 150B Neural Tokens
OmniMouse models brain activity, showing performance scales with data, not model size, unlike standard AI.
High-fidelity and Network-based Spatio-temporal Mathematical Models of Alzheimer's Disease Progression and their Validation Against PET-SUVR Imaging Data
This paper compares high-fidelity 3D and network-based spatio-temporal models for Alzheimer's disease progression, validated with PET-SUVR data.
The Umwelt Representation Hypothesis: Rethinking Universality
This paper introduces the Umwelt Representation Hypothesis, arguing that representational alignment in brains and ANNs stems from shared ecological constraints, not universal convergence.
Quantum-Like Models of Cognition and Decision Making: Open-Systems and Gorini--Kossakowski--Sudarshan--Lindblad Dynamics
This paper introduces a dynamical framework using GKSL equations to model cognitive processes, decision-making, and internal mental struggles.
How Much Data is Enough? The Zeta Law of Discoverability in Biomedical Data, featuring the enigmatic Riemann zeta function
This paper introduces a zeta-law scaling framework to predict when more biomedical data, better representations, or new modalities will accelerate scientific discovery.
Poisson Flow Model of Cortical Folding Pattern
Introduces a Poisson flow model to characterize cortical folding patterns, offering a new way to study subtle brain abnormalities in JME.
NeuroAI and Beyond: Bridging Between Advances in Neuroscience and ArtificialIntelligence
This paper outlines how NeuroAI, by integrating neuroscience principles, can overcome current AI limitations in interaction, learning, and efficiency.
Causality as a Minimum Energy Principle
This paper introduces a variational causal framework that models causality as energy flow, effectively capturing cyclic and higher-order dynamics in complex networks.
Untrained CNNs Match Backpropagation at V1: A Systematic RSA Comparison of Four Learning Rules Against Human fMRI
Untrained CNNs achieve V1/V2 alignment with human fMRI data comparable to backpropagation, highlighting architecture's dominant role in early visual processing.
Timescale Limits of Linear-Threshold Networks
This paper explores global stability in linear-threshold networks by analyzing fast and slow limits, revealing key stability mechanisms.
Role of chloride concentration in modulating seizure transitions in excitatory and inhibitory networks
This paper models how chloride concentration dynamics, specifically inhibitory synaptic conductance, control seizure initiation and progression through distinct stages.
Goxpyriment: A Go Framework for Behavioral and Cognitive Experiments
Goxpyriment is a new Go framework for behavioral and cognitive experiments, simplifying deployment with self-contained executables and ensuring precise timing.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.