Predictive Coding Theories of Cortical Function

Rao, Rajesh P.N.; Jiang, Linxing Preston

doi:10.1093/acrefore/9780190264086.013.328

Cited by 19 publications

(13 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…One interpretation of their functional relevance is that these timescales set the duration over which a neural circuit integrates its inputs 12 . In this interpretation, shorter timescales in sensory areas enable them to rapidly detect or faithfully track dynamic stimuli 13,14 . By contrast, prefrontal areas can utilize longer timescales to integrate information and improve the signal-to-noise ratio in short-term memory or decision-making computations 12,15 .…”

Section: Temporal Hierarchy Through Prediction Of Dynamicsmentioning

confidence: 99%

Dynamic Predictive Coding: A Model of Hierarchical Sequence Learning and Prediction in the Neocortex

Jiang

Rao

2022

Preprint

Self Cite

View full text Add to dashboard Cite

We introduce dynamic predictive coding, a new hierarchical model of spatiotemporal prediction and sequence learning in the cortex. The model assumes that higher cortical levels modulate the temporal dynamics of lower levels, correcting their predictions of dynamics using precision-weighted prediction errors. We tested this model using a two-level neural network, where the top-down modulation is implemented as a low-dimensional mixture of possible temporal dynamics. When trained on natural videos, the first-level neurons developed space-time receptive fields similar to those found in simple cells of the primary visual cortex. The second-level responses spanned longer timescales and showed more stability than those at the first level, mimicking temporal response hierarchies in the cortex. After adapting to a repeated visual sequence, the model displayed full recall of the sequence given only the beginning of the sequence, similar to sequence recall in the visual cortex. Our results suggest that sequence learning and temporal prediction in the cortex can be interpreted as dynamic predictive coding based on a hierarchical generative model of input sequences.

show abstract

Section: Temporal Hierarchy Through Prediction Of Dynamicsmentioning

confidence: 99%

Dynamic Predictive Coding: A Model of Hierarchical Sequence Learning and Prediction in the Neocortex

Jiang

Rao

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…A second motif that distinguishes predictive coding models from the hierarchical inference model presented here is the role of error coding units. Predictive coding postulates the existence of two distinct classes of neurons [64]. Neurons in the deeper 5/6 layers of the cortex are responsible for internal state representation while 2/3 layer neurons are error detecting neurons representing the difference of sensory input and top-down prediction.…”

Section: Discussionmentioning

confidence: 99%

Top-down perceptual inference shaping the activity of early visual cortex

Csikor,

Meszéna,

Orbán

2023

Preprint

View full text Add to dashboard Cite

Deep discriminative models provide remarkable insights into hierarchical processing in the brain by predicting neural activity along the visual pathway. However, these models differ from biological systems in their computational and architectural properties. Unlike biological systems, they require teaching signals for supervised learning. Moreover, they rely on feed-forward processing of stimuli, which contrasts with the extensive top-down connections in the ventral pathway. Here, we address both issues by developing a hierarchical deep generative model and show that it predicts an extensive set of experimental results in the primary and secondary visual cortices (V1 and V2). We show that the widely documented nonlinear sensitivity of V2 neurons to texture statistics is a consequence of learning a hierarchical representation of natural images. Further, we show that top-down influences are inherent to inference in hierarchical generative models, and explain neuronal responses to illusory contours and systematic modulations of noise correlations in V1.

show abstract

“…The key concepts, principles, and mechanisms that this version of PP relies on are gaining more and more theoretical and empirical support (see Clark, 2016 or Hohwy, 2020 for reviews). The implied approximate Bayesian nature of cognition is suggested by many empirical studies (Jiang & Rao, 2021; Millidge et al, 2021; Walsh et al, 2020).…”

Section: The Predictive Processing Framework In Cognitive Sciencementioning

confidence: 97%

Conceptual engineering, predictive processing, and a new implementation problem

Löhr

Michel

2023

Mind & Language

View full text Add to dashboard Cite

According to predictive processing, an increasingly influential paradigm in cognitive science, the function of the brain is to minimize the prediction error of its sensory input. Conceptual engineering is the practice of assessing and changing concepts or word meanings. We contribute to both strands of research by proposing the first cognitive account of conceptual engineering, using the predictive processing framework. Our model reveals a new kind of implementation problem as prediction errors are only minimized if enough agents embrace conceptual changes. This problem can be overcome by emphasizing the importance of social norms and conceptual pluralism.

show abstract

Predictive Coding Theories of Cortical Function

Cited by 19 publications

References 0 publications

Dynamic Predictive Coding: A Model of Hierarchical Sequence Learning and Prediction in the Neocortex

Dynamic Predictive Coding: A Model of Hierarchical Sequence Learning and Prediction in the Neocortex

Top-down perceptual inference shaping the activity of early visual cortex

Conceptual engineering, predictive processing, and a new implementation problem

Contact Info

Product

Resources

About