interpretabilityfactbullishModern ILP frameworks can learn complex, non-monotonic hypotheses that broaden modeling capabilities for real-world applicationsMachine Learning28 Jul 2026http://arxiv.org/abs/2607.21184v1