Variability, Bayesian Inference, and Sensorimotor Learning

By combining long-term manipulations of auditory feedback with mathematical models of sensorimotor learning, we reveal the computational principles underlying vocal learning.

Millisecond-scale motor coding precedes sensorimotor learning in songbirds
Leila May M. Pascual, Aanya Vusirikala, Ilya M. Nemenman, Samuel J. Sober, Michael Pasek. BioArXiv 2024.09.27.615500 (2024) [PDF] [DOI]

---

Chance, long tails, and inference in a non-Gaussian, Bayesian theory of vocal learning in songbirds
Zhou B, Hofmann D, Pinkoviezky I, Sober SJ, Nemenman I. Proceedings of the National Academy of Sciences 115(36):E8538-E8546 (2018) [PDF]

---

A simple computational principle predicts vocal adaptation dynamics across age and error size
Kelly CW, Sober SJ. Frontiers in Integrative Neuroscience doi: 10.3389/fnint.2014.00075 (2014) [PDF] [DOI]

---