Gated recurrence enables simple and accurate sequence prediction in stochastic, changing, and structured environments
- INSERM, CEA, Université Paris-Saclay, France
- CEA, Sorbonne Université, France
Abstract
From decision making to perception to language, predicting what is coming next is crucial. It is also challenging in stochastic, changing, and structured environments; yet the brain makes accurate predictions in many situations. What computational architecture could enable this feat? Bayesian inference makes optimal predictions but is prohibitively difficult to compute. Here, we show that a specific recurrent neural network architecture enables simple and accurate solutions in several environments. This architecture relies on three mechanisms: gating, lateral connections, and recurrent weight training. Like the optimal solution and the human brain, such networks develop internal representations of their changing environment (including estimates of the environment's latent variables and the precision of these estimates), leverage multiple levels of latent structure, and adapt their effective learning rate to changes without changing their connection weights. Being ubiquitous in the brain, gated recurrence could therefore serve as a generic building block to predict in real-life environments.
Data availability
This paper presents no experimental data. All synthetic data are available in the code repository at https://github.com/cedricfoucault/networks_for_sequence_prediction and archived on Zenodo with DOI: 10.5281/zenodo.5707498.
-
Networks for sequence predictionPublicly available at Github (https://github.com).
-
Networks for sequence predictionPublicly available at Zenodo (https://zenodo.org/).
Article and author information
Author details
Funding
Ecole normale superieure Paris-Saclay (PhD fellowship Contrat doctoral spécifique normalien"")
- Cédric Foucault
Agence Nationale de la Recherche (18-CE37-0010-01 CONFI LEARN"")
- Florent Meyniel
H2020 European Research Council (ERC StG 947105 NEURAL PROB"")
- Florent Meyniel
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Copyright
© 2021, Foucault & Meyniel
This article is distributed under the terms of the Creative Commons Attribution License permitting unrestricted use and redistribution provided that the original author and source are credited.
Metrics
-
- 1,271
- views
-
- 193
- downloads
-
- 7
- citations
Views, downloads and citations are aggregated across all versions of this paper published by eLife.
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Gated recurrence enables simple and accurate sequence prediction in stochastic, changing, and structured environments
eLife 10:e71801.
https://doi.org/10.7554/eLife.71801
Further reading
-
- Neuroscience
By influencing calcium homeostasis, local protein synthesis and the endoplasmic reticulum, a small protein called Rab10 emerges as a crucial cytoplasmic regulator of neuropeptide secretion.
-
- Neuroscience
The infralimbic cortex (IL) is essential for flexible behavioral responses to threatening environmental events. Reactive behaviors such as freezing or flight are adaptive in some contexts, but in others a strategic avoidance behavior may be more advantageous. IL has been implicated in avoidance, but the contribution of distinct IL neural subtypes with differing molecular identities and wiring patterns is poorly understood. Here, we study IL parvalbumin (PV) interneurons in mice as they engage in active avoidance behavior, a behavior in which mice must suppress freezing in order to move to safety. We find that activity in inhibitory PV neurons increases during movement to avoid the shock in this behavioral paradigm, and that PV activity during movement emerges after mice have experienced a single shock, prior to learning avoidance. PV neural activity does not change during movement toward cued rewards or during general locomotion in the open field, behavioral paradigms where freezing does not need to be suppressed to enable movement. Optogenetic suppression of PV neurons increases the duration of freezing and delays the onset of avoidance behavior, but does not affect movement toward rewards or general locomotion. These data provide evidence that IL PV neurons support strategic avoidance behavior by suppressing freezing.
