English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Machine learning algorithms have achieved superhuman performance in specificcomplex domains. Yet learning online from few examples and efficientlygeneralizing across domains remains elusive. In humans such learning proceedsvia declarative memory formation and is closely associated with consciousness.Predictive processing has been advanced as a principled Bayesian inferenceframework for understanding the cortex as implementing deep generativeperceptual models for both sensory data and action control. However, predictiveprocessing offers little direct insight into fast compositional learning or themystery of consciousness. Here we propose that through implementing onlinelearning by hierarchical binding of unpredicted inferences, a predictiveprocessing system may flexibly generalize in novel situations by formingworking memories for perceptions and actions from single examples, which canbecome short- and long-term declarative memories retrievable by associativerecall. We argue that the contents of such working memories are unified yetdifferentiated, can be maintained by selective attention and are consistentwith observations of masking, postdictive perceptual integration, and otherparadigm cases of consciousness research. We describe how the brain could haveevolved to use perceptual value prediction for reinforcement learning ofcomplex action policies simultaneously implementing multiple survival andreproduction strategies. 'Conscious experience' is how such a learning systemperceptually represents its own functioning, suggesting an answer to the metaproblem of consciousness. Our proposal naturally unifies feature binding,recurrent processing, and predictive processing with global workspace, and, toa lesser extent, the higher order theories of consciousness.

Consciousness is learning: predictive processing systems that learn by binding may perceive themselves as conscious