Computational Neuroscience
Computational models of the brain, neural coding, and brain-computer interfaces.
q-bio.NC · 115 papersRobust Evaluation of Neural Encoding Models via ground-truth approximation
Introduces CPA-PA, a robust evaluation framework for neural encoding models that approximates ground-truth neural activity, outperforming conventional metrics.
Working Memory in a Recurrent Spiking Neural Networks With Heterogeneous Synaptic Delays
Recurrent spiking neural networks with heterogeneous synaptic delays enable perfect working memory of temporal spike patterns.
Seeing the imagined: a latent functional alignment in visual imagery decoding from fMRI data
This paper adapts a SOTA fMRI decoder for visual perception to reconstruct imagined content using latent functional alignment and data augmentation.
Modeling of Self-sustained Neuron Population without External Stimulus
Hodgkin-Huxley neuron networks with STDP and stochasticity can sustain long-duration, sparse, irregular autonomous activity after brief initial stimulation.
From Brain Models to Executable Digital Twins: Execution Semantics and Neuro-Neuromorphic Systems
This survey introduces physically constrained executability to unify approaches for brain digital twins, focusing on execution semantics across diverse systems.
Attention to task structure for cognitive flexibility
This paper explores how environmental task structure and attention models influence cognitive flexibility, stability, and generalization in multi-task learning.
The illusory simplicity of the feedforward pass: evidence for the dynamical nature of stimulus encoding along the primate ventral stream
Early primate vision isn't a simple feedforward sweep; it's a dynamic spatiotemporal process encoding information in neural pattern changes.
Brain-DiT: A Universal Multi-state fMRI Foundation Model with Metadata-Conditioned Pretraining
Brain-DiT is a universal fMRI foundation model using metadata-conditioned diffusion pretraining across diverse brain states for generalized representations.
Machine learning approaches to uncover the neural mechanisms of motivated behaviour: from ADHD to individual differences in effort and reward sensitivity
This thesis uses machine learning on neuroimaging data to uncover neural mechanisms of motivated behavior, ADHD, and individual differences in effort/reward sensitivity.
Integrated information theory: the good, the bad and the misunderstood
This paper clarifies and critiques the Integrated Information Theory (IIT) of consciousness, addressing common misunderstandings and proposing future directions.
The Neurobiological Craving Signature (NCS) predicts social craving and responds to social isolation
The Neurobiological Craving Signature (NCS), previously linked to drug and food craving, also predicts social craving and responds to social isolation.
Probabilistic Prediction of Neural Dynamics via Autoregressive Flow Matching
Introduces Autoregressive Flow Matching (AFM) for probabilistic forecasting of neural dynamics, outperforming baselines on fMRI data.
Relaxing in Warped Spaces: Generalized Hierarchical and Modular Dynamical Neural Network
A novel hierarchical and modular dynamical neural network learns complex mappings and associates information by relaxing in warped spaces.
Astrocytic resource diffusion stabilizes persistent activity in neural fields
This paper introduces an astrocyte-neural field model where astrocytic resource diffusion stabilizes persistent neural activity, crucial for working memory.
The Rise and Fall of $G$ in AGI
This paper applies psychometric g-factor analysis to LLM benchmarks, finding a strong general intelligence (G) that is now specializing as models outsource reasoning.
The Fast Lane Hypothesis: Von Economo Neurons Implement a Biological Speed-Accuracy Tradeoff
The Fast Lane Hypothesis proposes Von Economo neurons (VENs) implement a biological speed-accuracy tradeoff for rapid social decision-making.
Meta-learning In-Context Enables Training-Free Cross Subject Brain Decoding
A meta-learning method allows training-free, cross-subject fMRI brain decoding by inferring individual neural patterns in-context, eliminating fine-tuning.
Neuromodulation supports robust rhythmic pattern transitions in degenerate central pattern generators with fixed connectivity
This paper presents a neuromodulation-based control architecture for robustly switching rhythmic patterns in degenerate neural networks with fixed connectivity.
The Cartesian Cut in Agentic AI
This paper introduces "Cartesian agency" in LLM-based agents, where a learned core is separated from an engineered runtime, discussing its pros, cons, and alternatives.
The Principle of Maximum Heterogeneity Optimises Productivity in Distributed Production Systems Across Biology, Economics, and Computing
This paper introduces the Principle of Maximum Heterogeneity, showing how diverse configurations optimize productivity across biological, economic, and computational systems.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.