Adaptive AI Tutor: Personalized Learning Path Simulator (2D)
2D knowledge-ring view of the same AI tutoring mechanic as the 3D version: a central tutor core fires practice beams at skill nodes arranged around it, choosing each target with item-response theory so the beam colour — green or red — auto-grades the attempt live.
This 2D companion runs the identical item-response-theory tutoring loop as the 3D version, laid flat as a knowledge ring instead of a sphere: twenty skill nodes sit at fixed difficulties around a central AI tutor core, and every practice round the core samples the unmastered skill closest to the learner's current ability estimate θ, fires a beam out to it, and auto-grades the attempt with a logistic probability curve. A correct answer turns the node green and nudges θ upward; a miss turns it red and pulls θ back, while every skill quietly forgets a slice of its mastery each second it isn't reviewed — forcing the tutor to loop back for spaced repetition exactly as a real adaptive-learning platform would.
2D knowledge-ring dashboard of an IRT-driven AI tutor: a central core fires practice beams at skill nodes chosen by selection weight ∝ e^(−a·(θ−b)²)·(1−mastery), grading each attempt with P(correct) = 1/(1+e^(−a·(θ−b))) and applying a continuous forgetting decay between rounds.
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install