Why the simplest explanation isn't always the best

Assistive sensory-motor perturbations influence learned neural representations

Task errors are used to learn and refine motor skills. We investigated how task assistance influences learned neural representations using Brain-Computer Interfaces (BCIs), which map neural activity into movement via a decoder. We analyzed motor cortex activity as monkeys practiced BCI with a decoder that adapted to improve or maintain performance over days. Over time, task-relevant information became concentrated in fewer neurons, unlike with fixed decoders. At the population level, task information also became largely confined to a few neural modes that accounted for an unexpectedly small fraction of the population variance. A neural network model suggests the adaptive decoders directly contribute to forming these more compact neural representations. Our findings show that assistive decoders manipulate error information used for long-term learning computations like credit assignment, which informs our understanding of motor learning and has implications for designing real-world BCIs.

Geometry Linked to Untangling Efficiency Reveals Structure and Computation in Neural Populations

From an eagle spotting a fish in shimmering water to a scientist extracting patterns from noisy data, many cognitive tasks require untangling overlapping signals. Neural circuits achieve this by transforming complex sensory inputs into distinct, separable representations that guide behavior. Data-visualization techniques convey the geometry of these transformations, and decoding approaches quantify performance efficiency. However, we lack a framework for linking these two key aspects. Here we address this gap by introducing a data-driven analysis framework, which we call Geometry Linked to Untangling Efficiency (GLUE) with manifold capacity theory, that links changes in the geometrical properties of neural activity patterns to representational untangling at the computational level. We applied GLUE to over seven neuroscience datasets-spanning multiple organisms, tasks, and recording techniques-and found that task-relevant representations untangle in many domains, including along the cortical hierarchy, through learning, and over the course of intrinsic neural dynamics. Furthermore, GLUE can characterize the underlying geometric mechanisms of representational untangling, and explain how it facilitates efficient and robust computation. Beyond neuroscience, GLUE provides a powerful framework for quantifying information organization in data-intensive fields such as structural genomics and interpretable AI, where analyzing high-dimensional representations remains a fundamental challenge.

DySCo: A general framework for dynamic functional connectivity

A crucial challenge in neuroscience involves characterising brain dynamics from high-dimensional brain recordings. Dynamic Functional Connectivity (dFC) is an analysis paradigm that aims to address this challenge. dFC consists of a time-varying matrix (dFC matrix) expressing how pairwise interactions across brain areas change over time. However, the main dFC approaches have been developed and applied mostly empirically, lacking a common theoretical framework and a clear view on the interpretation of the results derived from the dFC matrices. Moreover, the dFC community has not been using the most efficient algorithms to compute and process the matrices efficiently, which has prevented dFC from showing its full potential with high-dimensional datasets and/or real-time applications. In this paper, we introduce the Dynamic Symmetric Connectivity Matrix analysis framework (DySCo), with its associated repository. DySCo is a framework that presents the most commonly used dFC measures in a common language and implements them in a computationally efficient way. This allows the study of brain activity at different spatio-temporal scales, down to the voxel level. DySCo provides a single framework that allows to: (1) Use dFC as a tool to capture the spatio-temporal interaction patterns of data in a form that is easily translatable across different imaging modalities. (2) Provide a comprehensive set of measures to quantify the properties and evolution of dFC over time: the amount of connectivity, the similarity between matrices, and their informational complexity. By using and combining the DySCo measures it is possible to perform a full dFC analysis. (3) Leverage the Temporal Covariance EVD algorithm (TCEVD) to compute and store the eigenvectors and values of the dFC matrices, and then also compute the DySCo measures from the EVD. Developing the framework in the eigenvector space is orders of magnitude faster and more memory efficient than naïve algorithms in the matrix space, without loss of information. The methodology developed here is validated on both a synthetic dataset and a rest/N-back task experimental paradigm from the fMRI Human Connectome Project dataset. We show that all the proposed measures are sensitive to changes in brain configurations and consistent across time and subjects. To illustrate the computational efficiency of the DySCo toolbox, we performed the analysis at the voxel level, a task which is computationally demanding but easily afforded by the TCEVD.

Is Ockham's razor losing its edge? New perspectives on the principle of model parsimony

The preference for simple explanations, known as the parsimony principle, has long guided the development of scientific theories, hypotheses, and models. Yet recent years have seen a number of successes in employing highly complex models for scientific inquiry (e.g., for 3D protein folding or climate forecasting). In this paper, we reexamine the parsimony principle in light of these scientific and technological advancements. We review recent developments, including the surprising benefits of modeling with more parameters than data, the increasing appreciation of the context-sensitivity of data and misspecification of scientific models, and the development of new modeling tools. By integrating these insights, we reassess the utility of parsimony as a proxy for desirable model traits, such as predictive accuracy, interpretability, effectiveness in guiding new research, and resource efficiency. We conclude that more complex models are sometimes essential for scientific progress, and discuss the ways in which parsimony and complexity can play complementary roles in scientific modeling practice.

Cognitive abilities are associated with rapid dynamics of electrophysiological connectome states

Time-varying changes in whole-brain connectivity patterns, or connectome state dynamics, hold significant implications for cognition. However, connectome dynamics at fast (>1 Hz) timescales highly relevant to cognition are poorly understood due to the dominance of inherently slow fMRI in connectome studies. Here, we investigated the behavioral significance of rapid electrophysiological connectome dynamics using source-localized EEG connectomes during resting state (N = 926, 473 females). We focused on dynamic connectome features pertinent to individual differences, specifically those with established heritability: Fractional Occupancy (i.e., the overall duration spent in each recurrent connectome state) in beta and gamma bands and Transition Probability (i.e., the frequency of state switches) in theta, alpha, beta, and gamma bands. Canonical correlation analysis found a significant relationship between the heritable phenotypes of subsecond connectome dynamics and cognition. Specifically, principal components of Transition Probabilities in alpha (followed by theta and gamma bands) and a cognitive factor representing visuospatial processing (followed by verbal and auditory working memory) most notably contributed to the relationship. We conclude that rapid connectome state transitions shape individuals' cognitive abilities and traits. Such subsecond connectome dynamics may inform about behavioral function and dysfunction and serve as endophenotypes for cognitive abilities.

Brain-phenotype predictions of language and executive function can survive across diverse real-world data: Dataset shifts in developmental populations

Predictive modeling potentially increases the reproducibility and generalizability of neuroimaging brain-phenotype associations. Yet, the evaluation of a model in another dataset is underutilized. Among studies that undertake external validation, there is a notable lack of attention to generalization across dataset-specific idiosyncrasies (i.e., dataset shifts). Research settings, by design, remove the between-site variations that real-world and, eventually, clinical applications demand. Here, we rigorously test the ability of a range of predictive models to generalize across three diverse, unharmonized developmental samples: the Philadelphia Neurodevelopmental Cohort (n=1291), the Healthy Brain Network (n=1110), and the Human Connectome Project in Development (n=428). These datasets have high inter-dataset heterogeneity, encompassing substantial variations in age distribution, sex, racial and ethnic minority representation, recruitment geography, clinical symptom burdens, fMRI tasks, sequences, and behavioral measures. Through advanced methodological approaches, we demonstrate that reproducible and generalizable brain-behavior associations can be realized across diverse dataset features. Results indicate the potential of functional connectome-based predictive models to be robust despite substantial inter-dataset variability. Notably, for the HCPD and HBN datasets, the best predictions were not from training and testing in the same dataset (i.e., cross-validation) but across datasets. This result suggests that training on diverse data may improve prediction in specific cases. Overall, this work provides a critical foundation for future work evaluating the generalizability of brain-phenotype associations in real-world scenarios and clinical settings.

Molecular causality in the advent of foundation models

Correlation is not causation: this simple and uncontroversial statement has far-reaching implications. Defining and applying causality in biomedical research has posed significant challenges to the scientific community. In this perspective, we attempt to connect the partly disparate fields of systems biology, causal reasoning, and machine learning to inform future approaches in the field of systems biology and molecular medicine.

Brain-phenotype predictions can survive across diverse real-world data

Recent work suggests that machine learning models predicting psychiatric treatment outcomes based on clinical data may fail when applied to unharmonized samples. Neuroimaging predictive models offer the opportunity to incorporate neurobiological information, which may be more robust to dataset shifts. Yet, among the minority of neuroimaging studies that undertake any form of external validation, there is a notable lack of attention to generalization across dataset-specific idiosyncrasies. Research settings, by design, remove the between-site variations that real-world and, eventually, clinical applications demand. Here, we rigorously test the ability of a range of predictive models to generalize across three diverse, unharmonized samples: the Philadelphia Neurodevelopmental Cohort (n=1291), the Healthy Brain Network (n=1110), and the Human Connectome Project in Development (n=428). These datasets have high inter-dataset heterogeneity, encompassing substantial variations in age distribution, sex, racial and ethnic minority representation, recruitment geography, clinical symptom burdens, fMRI tasks, sequences, and behavioral measures. We demonstrate that reproducible and generalizable brain-behavior associations can be realized across diverse dataset features with sample sizes in the hundreds. Results indicate the potential of functional connectivity-based predictive models to be robust despite substantial inter-dataset variability. Notably, for the HCPD and HBN datasets, the best predictions were not from training and testing in the same dataset (i.e., cross-validation) but across datasets. This result suggests that training on diverse data may improve prediction in specific cases. Overall, this work provides a critical foundation for future work evaluating the generalizability of neuroimaging predictive models in real-world scenarios and clinical settings.