Decoding Dynamic Visual Experience from Calcium Imaging via Cell-Pattern-Aware Pretraining

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bae, Sangyoon, Azabou, Mehdi, Richards, Blake, Cha, Jiook
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910199763173376
author Bae, Sangyoon
Azabou, Mehdi
Richards, Blake
Cha, Jiook
author_facet Bae, Sangyoon
Azabou, Mehdi
Richards, Blake
Cha, Jiook
contents Neural recordings exhibit a distinctive form of heterogeneity rooted in differences in cell types, intrinsic circuit dynamics, and stochastic stimulus-response variability that goes beyond ordinary dataset variability, mixing statistically regular neurons with highly stochastic, stimulus-contingent ones within the same dataset. This heterogeneity poses a challenge for self-supervised learning (SSL) -- learnable statistical regularity -- thereby destabilizing representation learning and limiting reliable scaling. We introduce POYO-CAP (Cell-pattern Aware Pretraining), a biologically grounded hybrid pretraining strategy that first trains with masked reconstruction plus lightweight auxiliary supervision on statistically regular neurons -- identified via skewness and kurtosis -- and then fine-tunes on more stochastic populations. On the Allen Brain Observatory dataset, this curriculum yields 12--13\% relative improvements over from-scratch training and enables smooth, monotonic scaling with model size, whereas baselines trained on mixed populations plateau or destabilize. By making statistical predictability an explicit data-selection criterion, POYO-CAP turns neural heterogeneity into a scalable learning advantage for robust neural decoding.
format Preprint
id arxiv_https___arxiv_org_abs_2510_18516
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Decoding Dynamic Visual Experience from Calcium Imaging via Cell-Pattern-Aware Pretraining
Bae, Sangyoon
Azabou, Mehdi
Richards, Blake
Cha, Jiook
Neurons and Cognition
Machine Learning
Neural recordings exhibit a distinctive form of heterogeneity rooted in differences in cell types, intrinsic circuit dynamics, and stochastic stimulus-response variability that goes beyond ordinary dataset variability, mixing statistically regular neurons with highly stochastic, stimulus-contingent ones within the same dataset. This heterogeneity poses a challenge for self-supervised learning (SSL) -- learnable statistical regularity -- thereby destabilizing representation learning and limiting reliable scaling. We introduce POYO-CAP (Cell-pattern Aware Pretraining), a biologically grounded hybrid pretraining strategy that first trains with masked reconstruction plus lightweight auxiliary supervision on statistically regular neurons -- identified via skewness and kurtosis -- and then fine-tunes on more stochastic populations. On the Allen Brain Observatory dataset, this curriculum yields 12--13\% relative improvements over from-scratch training and enables smooth, monotonic scaling with model size, whereas baselines trained on mixed populations plateau or destabilize. By making statistical predictability an explicit data-selection criterion, POYO-CAP turns neural heterogeneity into a scalable learning advantage for robust neural decoding.
title Decoding Dynamic Visual Experience from Calcium Imaging via Cell-Pattern-Aware Pretraining
topic Neurons and Cognition
Machine Learning
url https://arxiv.org/abs/2510.18516