Every receptor binding, every sniff, and every oscillation in this simulation emerges from CRR (Coherence-Rupture-Regeneration), a temporal grammar for the Free Energy Principle. Three equations, one parameter, zero free parameters.
Olfaction is the brain's most prediction-dependent sense. The chemical world is noisy and turbulent, so the system invests more precision in its prior model than in sensory evidence alone.
Sniffing is active inference: a controlled sensory sampling act driven by predictions about what odors to expect, like a saccade or a whisker sweep.
Each olfactory receptor neuron expresses one receptor type from ~400 genes. Odorant binding is bistable: the molecule fits or it does not.
The swipe card model (Brookes et al. 2007) proposes that signal transduction requires shape fit (Z₂) AND vibrational frequency match (SO(2)): phonon-assisted inelastic electron tunnelling.
Piriform cortex shows theta (4.72 Hz), beta (18.98 Hz), and gamma (84.94 Hz). Gamma amplitude is modulated by theta phase, ONLY when odor is present. Gamma/theta = 18 = 9 × 2.
CRR predicts the resting rate from theta: period = 3 × 2π/θ = 6π/4.72 = 4.0 s = 15/min. Three SO(2) revolutions of the piriform theta pacemaker per breath. Zero free parameters.
B(C) = exp(C/Ω)·(C* − C) peaks at C* − Ω: 89.9% through each gamma cycle. Maximum olfactory sensitivity one Ω before rupture. The golden glow in the simulation marks this moment.
Parr, Pezzulo & Friston (2022) Active Inference. MIT Press.
Stern et al. (2022) Gamma in piriform cortex. PLOS Biol.
Brookes et al. (2007) Phonon tunnelling. PRL 98:038101.
Wu et al. (2024) 60 ms temporal resolution. Nat Hum Behav.
Shusterman et al. (2011) Precise sniff responses. Nat Neurosci.