What Is a Generative Model? Definitions, Disagreements, and Evaluation in Human Neuroimaging
Preprint
Abstract
The term generative model is widely used in human neuroimaging; however, its meaning is often left implicit. Prompted by observations and discussions at the 2025 Organization for Human Brain Mapping (OHBM) Annual Meeting, we surveyed members of the neuroimaging community to examine how generative models are defined, used, and evaluated in practice. Responses revealed some agreement on functional criteria— such as a model’s ability to simulate data—alongside marked disagreement about whether specific, widely used methods should be considered generative models. Evaluative priorities also varied across respondents, though out-of-sample generalization and interpretability were consistently emphasized. Rather than proposing a single definition, this perspective highlights the diversity of current usage and argues for greater clarity when the term is invoked.Keywords
This publication has 20 references indexed in Scilit:
- Generative artificial intelligence: a systematic review and applicationsMultimedia Tools and Applications, 2024
- Exploration, Inference, and Prediction in Neuroscience and BiomedicineTrends in Neurosciences, 2019
- Regression DCM for fMRINeuroImage, 2017
- A DCM for resting state fMRINeuroImage, 2014
- Simple group fMRI modeling and inferenceNeuroImage, 2009
- Modeling the hemodynamic response function in fMRI: Efficiency, bias and mis-modelingNeuroImage, 2009
- Dynamic causal modellingNeuroImage, 2003
- Predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effectsNature Neuroscience, 1999
- Statistical parametric maps in functional imaging: A general linear approachHuman Brain Mapping, 1994
- A learning algorithm for boltzmann machinesCognitive Science, 1985