mercurial.spectral.perception module

Bayesian perception as pattern completion (MERCURIAL D).

class mercurial.spectral.perception.BayesianPerception(hierarchy: ModalityEnergyHierarchy, prior_strength: float = 1.0, learning_rate: float = 0.1)

Bases: object

Implements perceptual inference via variational free energy minimization.

Methods

detect(pattern_intensity, ...)	Return True if pattern is consciously perceived.
detection_threshold(pattern_intensity, ...)	Θ_perc = Θ_0 * (1 + β_state ψ_obs) / (1 + γ_noise N_env) * (1 - S/S_crit)
dpr_perception(sensory_input, ...[, ...])	Use optimal DPR pathway for perception.
free_energy(sensory_input, predicted_pattern)	F_perc = ||sensory - predicted||^2 + β_prior ||predicted - prior||^2
infer_hidden_state(sensory_input[, prior, ...])	Find pattern that minimizes free energy (gradient descent).

__init__(hierarchy: ModalityEnergyHierarchy, prior_strength: float = 1.0, learning_rate: float = 0.1)

Parameters:

hierarchyModalityEnergyHierarchy

Energy hierarchy for modality thresholds.

prior_strengthfloat

β_prior for prior influence.

learning_ratefloat

Step size for gradient descent.

detect(pattern_intensity: float, observer_coherence: float, environmental_noise: float) → bool

Return True if pattern is consciously perceived.

detection_threshold(pattern_intensity: float, observer_coherence: float, environmental_noise: float) → float

Θ_perc = Θ_0 * (1 + β_state ψ_obs) / (1 + γ_noise N_env) * (1 - S/S_crit)

dpr_perception(sensory_input: ndarray, pattern_energy: float, pattern_coherence: float, environmental_noise: float, branch_similarity: float = 1.0) → ndarray

Use optimal DPR pathway for perception.

free_energy(sensory_input: ndarray, predicted_pattern: ndarray, prior_pattern: ndarray | None = None) → float

F_perc = ||sensory - predicted||^2 + β_prior ||predicted - prior||^2

infer_hidden_state(sensory_input: ndarray, prior: ndarray | None = None, max_iter: int = 20) → ndarray

Find pattern that minimizes free energy (gradient descent).