3D Lecture Hall: Belief Correction via Costly-Accurate Signals
3D Bayesian model of a lecture hall of students whose belief in a false explanation converges toward truth at a rate set by evidence quality and each student's own engagement, visualized as instanced seats shifting color as their log-odds update every semester.
A lecture hall of instanced student seats, each tracking its own belief in a false explanation as a Bayesian log-odds value. Every simulated evidence pulse from the podium nudges every seat's log-odds toward truth, scaled by evidence quality and that student's individual engagement, so the hall's color field desaturates from red toward gold at an uneven but statistically predictable pace over simulated semesters — a direct 3D read of the same drift-diffusion process driving the 2D belief-over-time chart in the companion simulation.
In a 3D lecture hall, student nodes visually converge toward the truth at individually adjustable rates.
3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install