analiza aiqbench
This study demonstrates that masked autoencoders can extract task-relevant structure from unlabeled spontaneous neural activity in the visual cortex of a blind participant. By pretraining on hours of resting-state data without supervision, the model achieved high accuracy in decoding perception tasks using linear probing, proving that self-supervised learning is viable for overcoming clinical neuroprosthetic data bottlenecks.
Masked Autoencoders Learn Perception-Relevant Representations from Resting State Neural Data
arXiv:2607.22615v1 Announce Type: cross Abstract: Clinical neuroprosthetics face a data bottleneck: labeled perception trials are scarce while hours of spontaneous neural activity are largely underutilized. Here, we test whether self-supervised learning can use these unlabeled datasets to improve perception decoding. We pretrained a masked autoencoder on 14.6 hours of spontaneous multiunit activity from an intracortical array in a blind participant's V1. The model captured interpretable brain structure without supervision: V1's spatial organization and perceptual state separation both emerged purely from its latent representations. To test these features, we used linear probing (logistic regression on the frozen latents) to measure performance on the data with stimulation. Perception decoding accuracy reached 84.1% on a general psychometric task. On the more difficult threshold-level task, accuracy reached 64.0%. This work shows that spontaneous cortical activity is not noise; it contains rich, task-relevant structure. Unsupervised pretraining on this data is a promising strategy to improve neural decoding.
Abstract:Clinical neuroprosthetics face a data bottleneck: labeled perception trials are scarce while hours of spontaneous neural activity are largely underutilized. Here, we test whether self-supervised learning can use these unlabeled datasets to improve perception decoding. We pretrained a masked autoencoder on 14.6 hours of spontaneous multiunit activity from an intracortical array in a blind participant's V1. The model captured interpretable brain structure without supervision: V1's spatial organization and perceptual state separation both emerged purely from its latent representations. To test these features, we used linear probing (logistic regression on the frozen latents) to measure performance on the data with stimulation. Perception decoding accuracy reached 84.1% on a general psychometric task. On the more difficult threshold-level task, accuracy reached 64.0%. This work shows that spontaneous cortical activity is not noise; it contains rich, task-relevant structure. Unsupervised pretraining on this data is a promising strategy to improve neural decoding.
6 papers, 4 figures. NeuroAI Workshop, AAAI 2026
Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
arXiv:2607.22615 [q-bio.NC]
(or arXiv:2607.22615v1 [q-bio.NC] for this version)
https://doi.org/10.48550/arXiv.2607.22615 arXiv-issued DOI via DataCite (pending registration)
Text integral sub licența deschisă a publicației · 185 cuvinte
Subiecte conexe
J++ Lens: Jacobian Filtering Enables More Faithful Workspace Lenses
LessWrong · 8 oct. 2026
PAIR: Bridging Perception and Action in Vision-Language-Action Models
arXiv · 8 oct. 2026
The J-Space Debate, Agent Swarms, and Pacing Frontier AI - Digital Minds Newsletter #4
LessWrong · 18 sept. 2026
Reading the Whole Heart: Latent-Attention Masked Autoencoders for Multimodal Cardiac Representation Learning
arXiv · 14 sept. 2026