BrainPFN: Amortized Brain-Behaviour Prediction via Prior-Data Fitted Networks
Abstract
The standard approach to predicting behavioral phenotypes from functional MRI is represented by Kernel Ridge Regression (KRR) applied on functional connectivity (FC) matrices. However, this approach is limited: separate models must be fit and tuned for every score and cohort, they do not transfer across datasets, and accuracy degrades sharply when only a handful of subjects are available, as is typical of clinical or cognitive studies. Despite an increasing interest in developing brain foundation models, none has consistently outperformed KRR on behaviour prediction; most reported gains come from easier targets such as age, sex, or diagnosis rather than continuous behavioural scores. We introduce BrainPFN, a Prior Data Fitted Network for brain behaviour regression. BrainPFN is meta-trained on a domain-specific prior built from ~7k real FC matrices, from which we sample one billion synthetic training samples. At inference, a labelled context is simply passed through the network in a single forward pass, with no per-score or per-cohort refitting required. Across four independent cohorts, BrainPFN beats KRR on small sample prediction, and leads recent foundation models pre-trained at four to nine times larger scale. Our meta-training pool, aggregated from open repositories, is a fraction of the size of the corpora behind competing models, most of which sit behind restrictive data use agreements such as UK Biobank. We release this pool publicly alongside code and checkpoints, showing that a modest, openly available pool can match or beat models trained at Biobank scale.
Related
- Protein Language Model-Conditioned Graph Neural Networks for Multitask GPCR Ligand Activity Prediction
- Personalized prediction of response to selective serotonin reuptake inhibitors (SSRIs) treatment for major depression disorder using deep learning networks from connectivity maps of EEG
- Covariate-aware genomic prediction of blood metabolite profiles using multi-task neural networks
- Evaluating multi-season occupancy models with autocorrelation fitted to heterogeneous datasets
- Whole-brain modeling of dynamic causal circuits in human cognition using amortized variational inference
BibTeX
@misc{barbano2026brainpfn,
title = {{BrainPFN: Amortized Brain-Behaviour Prediction via Prior-Data Fitted Networks}},
author = {Carlo Alberto Barbano and Demian Wassermann},
howpublished = {HAL},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2611.06934},
eprint = {2611.06934},
archivePrefix = {arXiv},
url = {https://inria.hal.science/hal-05767107v1},
note = {CC BY-NC 4.0},
}
RIS
TY - UNPB
TI - BrainPFN: Amortized Brain-Behaviour Prediction via Prior-Data Fitted Networks
AU - Carlo Alberto Barbano
AU - Demian Wassermann
PY - 2026
DA - 2026/09/28
PB - HAL
DO - 10.48550/arxiv.2611.06934
UR - https://inria.hal.science/hal-05767107v1
AB - The standard approach to predicting behavioral phenotypes from functional MRI is represented by Kernel Ridge Regression (KRR) applied on functional connectivity (FC) matrices. However, this approach is limited: separate models must be fit and tuned for every score and cohort, they do not transfer across datasets, and accuracy degrades sharply when only a handful of subjects are available, as is typical of clinical or cognitive studies. Despite an increasing interest in developing brain foundation models, none has consistently outperformed KRR on behaviour prediction; most reported gains come from easier targets such as age, sex, or diagnosis rather than continuous behavioural scores. We introduce BrainPFN, a Prior Data Fitted Network for brain behaviour regression. BrainPFN is meta-trained on a domain-specific prior built from ~7k real FC matrices, from which we sample one billion synthetic training samples. At inference, a labelled context is simply passed through the network in a single forward pass, with no per-score or per-cohort refitting required. Across four independent cohorts, BrainPFN beats KRR on small sample prediction, and leads recent foundation models pre-trained at four to nine times larger scale. Our meta-training pool, aggregated from open repositories, is a fraction of the size of the corpora behind competing models, most of which sit behind restrictive data use agreements such as UK Biobank. We release this pool publicly alongside code and checkpoints, showing that a modest, openly available pool can match or beat models trained at Biobank scale.
ER -