Hierarchical Gaussian Filter

Uses

How it works

Highlighted claims

• The paper expresses deep predictive coding networks as generalised HGFs to restore precision-weighted message passing and replace iterative inference. — Closed-form predictive coding via hierarchical Gaussian filters
• The HGF is a Bayesian model in which agents invert a generative model whose hidden states evolve as coupled random walks. — Closed-form predictive coding via hierarchical Gaussian filters
• Approximate HGF inversion yields closed-form one-step updates for posterior means and precisions. — Closed-form predictive coding via hierarchical Gaussian filters
• The HGF formulation reaches an analytic stationary point in one forward-backward sweep without iterating to equilibrium. — Closed-form predictive coding via hierarchical Gaussian filters