Abstract
Bayesian inference has emerged as a general framework that captures how organisms make decisions under uncertainty. Recent experimental findings reveal disparate mechanisms for how the brain generates behaviors predicted by normative Bayesian theories. Here, we identify two broad classes of neural implementations for Bayesian inference: a modular class, where each probabilistic component of Bayesian computation is independently encoded and a transform class, where uncertain measurements are converted to Bayesian estimates through latent processes. Many recent experimental neuroscience findings studying probabilistic inference broadly fall into these classes. We identify potential avenues for synthesis across these two classes and the disparities that, at present, cannot be reconciled. We conclude that to distinguish among implementation hypotheses for Bayesian inference, we require greater engagement among theoretical and experimental neuroscientists in an effort that spans different scales of analysis, circuits, tasks, and species.
Original language | English |
---|---|
Pages (from-to) | 121-129 |
Number of pages | 9 |
Journal | Current Opinion in Neurobiology |
Volume | 70 |
Early online date | 19 Oct 2021 |
DOIs | |
Publication status | Published - Oct 2021 |
Bibliographical note
Funding Information:The authors thank Joshua Dudman and Matthew Botvinick for inspiring discussions. The authors are grateful to Mehrdad Jazayeri and Seth Egger for their comments on the article. H.S. is supported by the Center for Sensorimotor Neural Engineering and the NARSAD young investigator grant (2020) by the Brain and Behavior Research Foundation . D.N. is supported by the Vidi grant ( 21-15016 ) from the Netherlands Organization for Scientific Research and by the EU Marie Sklodowska Curie reintegration grant (PredOpt 796577 ).
Publisher Copyright:
© 2021 The Author(s)