Category:
Foundational Concepts
In Active Inference, the generative model definition is: an agent’s internal ‘reality simulator.’ As detailed in the MIT Press text Active Inference, it is a probabilistic architecture that allows an agent to predict sensory inputs based on its internal beliefs.
Unlike standard AI which reacts to data, an agent with a generative model can imagine ‘what would happen if…’ scenarios. This allows it to act with purpose to minimize future surprise.
Key Insight: This model isn’t just a map; it is a dynamic tool for perception and action.
For a technical deep dive into how these maps function in autonomous systems, explore our guide on Active Inference World Models.
