Problem Solving as Probabilistic Inference with Subgoaling: Explaining Human Successes and Pitfalls in the Tower of Hanoi
Fig 3
Choice between two paths (w1 vs. w2) from S11 to S13 that require the same number of steps but traverse a different number of bottlenecks (1 for w1 and 2 for w2. Choice is performed using a probabilistic version of the drift diffusion model [72] introduced in the Section on ‘Methods’, in which particles (of the particle filtering algorithm) that reach the goal “vote” for the specific strategy they followed.