A Generalist Intracortical Motor Decoder
Joel Ye, Fabio Rizzoglio, Xuan Ma, Adam Smoulder, Hongwei Mao, Gary H Blumenthal, William Hockeimer, Nicolas Guazzelli Kunigk
Abstract
Mapping the relationship between neural activity and motor behavior is a central aim of sensorimotor neuroscience and neurotechnology. While most progress to this end has relied on restricting complexity, the advent of foundation models instead proposes integrating a breadth of data as an alternate avenue for broadly advancing downstream modeling. We quantify this premise for motor decoding from intracortical microelectrode data, pretraining an autoregressive Transformer on 2000 hours of neural population spiking activity paired with diverse motor covariates from over 30 monkeys and humans. The resulting model is broadly useful, benefiting decoding on 8 downstream decoding tasks and generalizing to a variety of neural distribution shifts. However, we also highlight that scaling autoregressive Transformers seems unlikely to resolve limitations stemming from sensor variability and output stereotypy in neural datasets.
BibTeX
@inproceedings{
ye2025a,
title={A Generalist Intracortical Motor Decoder},
author={Joel Ye and Fabio Rizzoglio and Xuan Ma and Adam Smoulder and Hongwei Mao and Gary H Blumenthal and William Hockeimer and Nicolas Guazzelli Kunigk and Dalton D. Moore and Patrick J. Marino and Raeed H. Chowdhury and J. Patrick Mayo and Aaron Batista and Steven Chase and Michael L Boninger and Charles M. Greenspon and Andrew B. Schwartz and Nicholas G. Hatsopoulos and Lee E. Miller and Kristofer Bouchard and Jennifer L Collinger and Leila Wehbe and Robert Gaunt},
booktitle={The Thirty-ninth Annual Conference on Neural Information Processing Systems},
year={2025},
url={https://openreview.net/forum?id=utXSSdD9mt}
}