Neural prediction decorrelation reveals that adversarial robustness substantially improves DNN prediction accuracy across the entire human auditory cortex
- 1University of Rochester Medical Center
Abstract
Sensory neuroscientists seek to model the neural computations that encode complex stimuli. Distinct encoding models often make similar predictions for natural stimuli such as speech, posing a challenge for model comparison. We developed a method to synthesize stimuli that decorrelate model predictions across a neural population, termed neural prediction decorrelation (NPD). Using fMRI responses to NPD sounds, we compared standard and adversarially robust deep neural network models of human auditory cortex. Prediction accuracy for NPD sounds was substantially better for the adversarially robust model in every region tested, an effect completely masked with natural sounds. Population responses to natural and synthesized NPD sounds shared an interpretable low-dimensional organization that was reproduced by the robust encoding model. NPD provides a general approach for comparing encoding models and reveals that adversarial robustness expands the predictive power of DNNs beyond natural stimuli, which is likely critical for targeting population activity through stimulus synthesis.
Related
- Pep-PU-GAN: Positive-Unlabeled Adversarial Learning for Peptide Function Prediction
- Hi-cGAN: Prediction of Hi-C interaction matrices with conditional generative adversarial networks
- In silico prediction of metabolic trait robustness in microbial cells
- Adversarial robustness does not entail human-aligned categorization
- Adversarial robustness of deep learning-based thyroid nodule segmentation in ultrasound
BibTeX
@misc{skrill2026neural,
title = {{Neural prediction decorrelation reveals that adversarial robustness substantially improves DNN prediction accuracy across the entire human auditory cortex}},
author = {D. Skrill and J. Feather and S. V. Norman-Haignere},
howpublished = {bioRxiv},
year = {2026},
month = aug,
doi = {10.64898/2026.08.05.743059},
url = {https://www.biorxiv.org/content/10.64898/2026.08.05.743059v2},
note = {CC BY},
}
RIS
TY - UNPB
TI - Neural prediction decorrelation reveals that adversarial robustness substantially improves DNN prediction accuracy across the entire human auditory cortex
AU - D. Skrill
AU - J. Feather
AU - S. V. Norman-Haignere
PY - 2026
DA - 2026/08/05
PB - bioRxiv
DO - 10.64898/2026.08.05.743059
UR - https://www.biorxiv.org/content/10.64898/2026.08.05.743059v2
AB - Sensory neuroscientists seek to model the neural computations that encode complex stimuli. Distinct encoding models often make similar predictions for natural stimuli such as speech, posing a challenge for model comparison. We developed a method to synthesize stimuli that decorrelate model predictions across a neural population, termed neural prediction decorrelation (NPD). Using fMRI responses to NPD sounds, we compared standard and adversarially robust deep neural network models of human auditory cortex. Prediction accuracy for NPD sounds was substantially better for the adversarially robust model in every region tested, an effect completely masked with natural sounds. Population responses to natural and synthesized NPD sounds shared an interpretable low-dimensional organization that was reproduced by the robust encoding model. NPD provides a general approach for comparing encoding models and reveals that adversarial robustness expands the predictive power of DNNs beyond natural stimuli, which is likely critical for targeting population activity through stimulus synthesis.
ER -