Retrieving "Generative Model" from the archives

Cross-reference notes under review

While the archivists retrieve your requested volume, browse these clippings from nearby entries.

1. Predictive Coding

   Linked via "generative model"

   The core mathematical structure of Predictive Coding is often framed within the principles of active inference, which seeks to minimize variational free energy ($F$). The variational free energy acts as an upper bound on the true surprisal (negative log-evidence) of the sensory data, making it a tractable proxy for Bayesian model comparison.
   The generative model ($P(\mathbf{x}, \mathbf{s})$) spec…

2. Predictive Coding

   Linked via "generative model"

   Precision Weighting and Attention
   The impact of a prediction error is not uniform; it is weighted by its precision$(\Pi(\epsilon))$. Precision is often equated with attentional allocation. High precision means the brain trusts that signal highly, leading to rapid updating of the generative model.
   $$