I'm curious, if it turns out that a simple rule-based algorithm exists, specifically tailored to solve (only!) ARC style problems, without generalization, would that still qualify for the reward?