The geometry of abstraction in hippocampus and prefrontal cortex

By S. Bernardi, Marcus K. Benna, Mattia Rigotti, C. Daniel Salzman. 2020

432 citations recorded by the source corpus.

SUMMARY The curse of dimensionality plagues models of reinforcement learning and decision-making. The process of abstraction solves this by constructing variables describing features shared by different instances, reducing dimensionality and enabling generalization in novel situations. Here we characterized neural representations in monkeys performing a task described by different hidden and explicit variables. Abstraction was defined operationally using the generalization performance of neural decoders across task conditions not used for training, which requires a particular geometry of neural representations. Neural ensembles in prefrontal cortex, hippocampus, and simulated neural networks simultaneously represented multiple variables in a geometry reflecting abstraction but that still allowed a linear classifier to decode a large number of other variables (high shattering dimensionality). Furthermore, this geometry changed in relation to task events and performance. These findings elucidate how the brain and artificial systems represent variables in an abstract format while preserving the advantages conferred by high shattering dimensionality.

Subfields: Artificial Intelligence, Cognitive Neuroscience

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Cognitive Neuroscience

Citations in the corpus

Orthogonal representations for robust context-dependent task performance in brains and neural networks (2022, neutral): “Another recent paper has emphasized that the neural geometry for distinct tasks in macaque PFC can become aligned along parallel manifolds, with representations for common action or outcome associations aligned in neural space (Bernardi et al., 2020).”
Low-dimensional dynamics for working memory and time encoding (2020, result): “S13), in agreement with other studies (40, 57).”
Human Replay Spontaneously Reorganizes Experience (2019, background): “Factorized representations are powerful, as they allow components to be recombined in many more ways than were experienced (Behrens et al., 2018; Bernardi et al., 2018; Higgins et al., 2017b).”
Interpreting neural computations by examining intrinsic and embedding dimensionality of neural activity (2021, background): “It is also possible to organize information embedding such that certain linear projection reflect information in more abstract form and therefore enable generalization, while others enable finer discrimination [15].”
Neural interactions in the human frontal cortex dissociate reward and punishment learning (2024, result): “This view is in line with recent literature showing that prefrontal cortex contains single neurons exhibiting mixed selectivity for multiple task variables (Meyers et al., 2008; Rigotti et al., 2013; Stokes et al., 2013; Panzeri et al., 2015; Parthasarathy et al., 2017; Bernardi et al., 2020).”