Linear NeuroPaint: An Efficient Baseline for Cross-Session Neural Alignment and Inpainting
- 1Columbia University
Abstract
Large-scale Neuropixels recordings sample many brain areas, but no single experimental session records every area of interest. NeuroPaint, a transformer-based autoencoder, integrates such recordings by aligning area-specific latent dynamics across sessions and "inpainting" activity in unrecorded areas. Could a linear model, which would be computationally efficient and potentially more interpretable, suffice to achieve this? To address this question, we introduce Linear NeuroPaint, a linear variant of NeuroPaint, and evaluate it on two multi-area Neuropixels datasets. Linear NeuroPaint achieves strong predictive performance on recorded areas, however, it is less accurate than nonlinear models at cross-area inpainting, despite extensive hyperparameter optimization. Moreover, its parameters are not readily interpretable: incomplete cross-session alignment leads parameter-defined neuron clusters to exhibit heterogeneous response profiles. Together, these results show that linear transformations are insufficient for cross-session alignment and cross-area prediction. Nevertheless, Linear NeuroPaint provides a computationally efficient baseline for benchmarking multi-session neural models.
Related
- DCSMAP: A Hybrid Linear-Pangenome Framework for Accurate and Efficient Read Alignment
- iTBS and cTBS to an individualized dlPFC target produce differential working memory improvement and session-dependent putamen deactivation
- Comparative Long-Term Immunogenicity and Boosting of 3-Session Versus 4-Session Intradermal Rabies Post-Exposure Prophylaxis Regimens in Cambodia and Madagascar: The RESIST-3 Study
- Repeated within-session rule switching in the marmoset: a paradigm for tracking the dynamics of cognitive flexibility
- Acute neuromuscular and kinematic responses to a single Feldenkrais awareness through movement session.
BibTeX
@misc{yao2026linear,
title = {{Linear NeuroPaint: An Efficient Baseline for Cross-Session Neural Alignment and Inpainting}},
author = {Y. J. Yao and J. Xia and L. Paninski and K. D. Miller},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.03.756082},
url = {https://www.biorxiv.org/content/10.64898/2026.10.03.756082v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Linear NeuroPaint: An Efficient Baseline for Cross-Session Neural Alignment and Inpainting
AU - Y. J. Yao
AU - J. Xia
AU - L. Paninski
AU - K. D. Miller
PY - 2026
DA - 2026/10/03
PB - bioRxiv
DO - 10.64898/2026.10.03.756082
UR - https://www.biorxiv.org/content/10.64898/2026.10.03.756082v1
AB - Large-scale Neuropixels recordings sample many brain areas, but no single experimental session records every area of interest. NeuroPaint, a transformer-based autoencoder, integrates such recordings by aligning area-specific latent dynamics across sessions and "inpainting" activity in unrecorded areas. Could a linear model, which would be computationally efficient and potentially more interpretable, suffice to achieve this? To address this question, we introduce Linear NeuroPaint, a linear variant of NeuroPaint, and evaluate it on two multi-area Neuropixels datasets. Linear NeuroPaint achieves strong predictive performance on recorded areas, however, it is less accurate than nonlinear models at cross-area inpainting, despite extensive hyperparameter optimization. Moreover, its parameters are not readily interpretable: incomplete cross-session alignment leads parameter-defined neuron clusters to exhibit heterogeneous response profiles. Together, these results show that linear transformations are insufficient for cross-session alignment and cross-area prediction. Nevertheless, Linear NeuroPaint provides a computationally efficient baseline for benchmarking multi-session neural models.
ER -