Bayesian Computation through Cortical Latent Dynamics
Statistical regularities in the environment create prior beliefs that we rely on to optimize our behavior when sensory information is uncertain. Bayesian theory formalizes how prior beliefs can be leveraged and has had a major impact on models of perception, sensorimotor function, and cognition. How...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Elsevier BV
2021
|
| Online Access: | https://hdl.handle.net/1721.1/130521 |
| _version_ | 1826199304280211456 |
|---|---|
| author | Sohn, Hansem Narain, Devika Jazayeri, Mehrdad |
| author2 | Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences |
| author_facet | Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences Sohn, Hansem Narain, Devika Jazayeri, Mehrdad |
| author_sort | Sohn, Hansem |
| collection | MIT |
| description | Statistical regularities in the environment create prior beliefs that we rely on to optimize our behavior when sensory information is uncertain. Bayesian theory formalizes how prior beliefs can be leveraged and has had a major impact on models of perception, sensorimotor function, and cognition. However, it is not known how recurrent interactions among neurons mediate Bayesian integration. By using a time-interval reproduction task in monkeys, we found that prior statistics warp neural representations in the frontal cortex, allowing the mapping of sensory inputs to motor outputs to incorporate prior statistics in accordance with Bayesian inference. Analysis of recurrent neural network models performing the task revealed that this warping was enabled by a low-dimensional curved manifold and allowed us to further probe the potential causal underpinnings of this computational strategy. These results uncover a simple and general principle whereby prior beliefs exert their influence on behavior by sculpting cortical latent dynamics. Sohn et al. found that prior beliefs warp neural representations in the frontal cortex. This warping provides a substrate for the optimal integration of prior beliefs with sensory evidence during sensorimotor behavior. |
| first_indexed | 2024-09-23T11:18:03Z |
| format | Article |
| id | mit-1721.1/130521 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T11:18:03Z |
| publishDate | 2021 |
| publisher | Elsevier BV |
| record_format | dspace |
| spelling | mit-1721.1/1305212022-09-27T18:33:25Z Bayesian Computation through Cortical Latent Dynamics Sohn, Hansem Narain, Devika Jazayeri, Mehrdad Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences Statistical regularities in the environment create prior beliefs that we rely on to optimize our behavior when sensory information is uncertain. Bayesian theory formalizes how prior beliefs can be leveraged and has had a major impact on models of perception, sensorimotor function, and cognition. However, it is not known how recurrent interactions among neurons mediate Bayesian integration. By using a time-interval reproduction task in monkeys, we found that prior statistics warp neural representations in the frontal cortex, allowing the mapping of sensory inputs to motor outputs to incorporate prior statistics in accordance with Bayesian inference. Analysis of recurrent neural network models performing the task revealed that this warping was enabled by a low-dimensional curved manifold and allowed us to further probe the potential causal underpinnings of this computational strategy. These results uncover a simple and general principle whereby prior beliefs exert their influence on behavior by sculpting cortical latent dynamics. Sohn et al. found that prior beliefs warp neural representations in the frontal cortex. This warping provides a substrate for the optimal integration of prior beliefs with sensory evidence during sensorimotor behavior. Netherlands Organization for Scientific Research (Rubicon Grant 446–14-008) Marie Sklodowska Curie Reintegration Grant (PredOpt 796577) National Institutes of Health (U.S.) ( (Grant NINDS-NS078127) 2021-04-26T12:49:07Z 2021-04-26T12:49:07Z 2019-09 2021-04-06T18:24:53Z Article http://purl.org/eprint/type/JournalArticle 0896-6273 https://hdl.handle.net/1721.1/130521 Sohn, Hansem et al. “Bayesian Computation through Cortical Latent Dynamics.” Neuron, 103, 5 (September 2019): 934–947.e5 © 2019 The Author(s) en 10.1016/J.NEURON.2019.06.012 Neuron Creative Commons Attribution-NonCommercial-NoDerivs License http://creativecommons.org/licenses/by-nc-nd/4.0/ application/pdf Elsevier BV PMC |
| spellingShingle | Sohn, Hansem Narain, Devika Jazayeri, Mehrdad Bayesian Computation through Cortical Latent Dynamics |
| title | Bayesian Computation through Cortical Latent Dynamics |
| title_full | Bayesian Computation through Cortical Latent Dynamics |
| title_fullStr | Bayesian Computation through Cortical Latent Dynamics |
| title_full_unstemmed | Bayesian Computation through Cortical Latent Dynamics |
| title_short | Bayesian Computation through Cortical Latent Dynamics |
| title_sort | bayesian computation through cortical latent dynamics |
| url | https://hdl.handle.net/1721.1/130521 |
| work_keys_str_mv | AT sohnhansem bayesiancomputationthroughcorticallatentdynamics AT naraindevika bayesiancomputationthroughcorticallatentdynamics AT jazayerimehrdad bayesiancomputationthroughcorticallatentdynamics |