Free Energy Principle
Free Energy Principle (FEP) is an idea that suggests that brain minimizes a quantity called Free Energy. Free Energy measures the difference between an organism's internal prediction of the world and the actual sensory input i.e. the "surprise".
FEP involves forming beliefs about the world and updating them as per sensory input. This similar to Bayesian models, but bayesian models account for beliefs and perceptions but not action.
Organisms achieve this free energy minimization through Active Inference. Active Inference accounts for actions too.
- An organism can minimize the free energy in one of two ways:
- Revise the prediction i.e. update the world model, or
- Act in a way to reduce its prediction error signals
So when we act, we are confirming our hypothesis, our fantasy, through actions
If I infer that I am touching my nose with my left index finger, but my proprioceptors tell me my arm is hanging at my side, I can minimize my brain’s raging prediction-error signals by raising that arm up and pressing a digit to the middle of my face. [1]
References:
- [1] The Genius Neuroscientist Who Might Hold the Key to True AI [wired.com]
See Also:
- https://rxinfer.com/ A Julia Package for Real-Time Bayesian Inference
- Neuroscience meets RL - AXIOM