Predictive Processing Flashcards
Predictive processing CLAIM that?
Predictive processing CLAIM that perceptual inference is done by following Bayes’ rule to infer causes of sensory input.
What is the belief that the brain is in predictive processing?
the brain is a prediction machine, which constantly generates models to anticipate sensory input, and uses these models or predictions to guide perception and action.
Predictive processin: Exploring the mind, by….
Exploring the brain
What is the brains overall goal in predictive processing?
The brain is only concerned to minimize prediction error
What is Bayes’ rule?
A normative tool/equation which can be used to predict outcomes and how likely a given event etc. are to happen based on a range of parameters.
What is prior probability?
Prior Probability (P(A)): This represents the brain’s prior belief or expectation about A before receiving the sensory input B. In this case, it is the probability of seeing a refrigerator in the kitchen based on past experience.
What is likelihood (Bayes’ rule)?
Likelihood (P(B∣A)): This is the probability of receiving the sensory input B given that the hypothesis A is true. It measures how likely the observed evidence is if the prediction is correct.
What is the posteriori probability?
Posterior Probability (P(A∣B)): After observing the evidence B, the brain updates its belief about the hypothesis A. This is the updated probability of A given the new evidence B.
Explain the process of Bayes’ rule using the terms: Prediction Generation, Prediction Error and Update beliefs
Prediction Generation: The brain generates predictions (A) about what it expects to perceive based on its internal model.
Prediction Error: The brain receives actual sensory input (B) and compares it to its predictions. The discrepancy between the predicted input and the actual input is the prediction error.
Update Beliefs: Using Bayes’ theorem, the brain updates its beliefs about the world. It adjusts the prior beliefs (P(A)) with the new evidence (B) to form the posterior beliefs (P(A∣B)).
What is the binocular rivavlvry experiment?
One eye sees a vertical grating and the other sees a horizontal grating, the brain has to choose between these two conflicting interpretations.
What is Alternation of Dominance? (Binocular rivalry)
Alternation of Dominance: Instead of integrating the conflicting inputs into a single coherent perception, the brain alternates between the two images. This can be seen as a process of the brain updating its hypothesis over time to minimize prediction error. When the prediction error for one image becomes too high, the brain switches to the other image.
What is perceptual inference?
Our own internal model of the world is updated based on incoming sensory input.
Is continuous, the brain tries to minimize the difference between its predictions and the actual sensory data.
What is active inference?
- We sample sensory input in the light of the model we hold about the world.
- Active inference involves the actions taken by an organism to sample sensory input in a way that aligns with its internal models.
- How the brain not only passively receives information but also actively seeks out information that confirms or disconfirms its predictions.
- Model-Based Action: The brain uses its internal models to decide on actions that will reduce uncertainty or prediction errors.
Explain the prediction error minimisation model
1) Build model of the world, H, based on sensory input
2) Evaluate H by taking action, thereby getting a new O,
using that to update the probability of H
3) Keep taking action to minimise the error on H to build
confidence in H
4) If new O indicates that H’ is more probable, abandon
H, and evaluate H’ by taking action
What is Predictive processing’s answer to the symbol grounding problem?
: Rather than seeking an external supervisor or programmer to provide truth, the predictive processing framework posits that the truth is embedded in the interaction between the brain’s predictions and the actual sensory input.
Predictive processing provides a solution to the symbol grounding problem by ensuring that symbols are not arbitrarily assigned meanings but are directly grounded in the sensory and motor interactions of the organism with the world.
