Towards a Solution to Bongard Problems: A Causal Approach

To date, Bongard Problems (BP) remain one of the few fortresses of AI history yet to be raided by the powerful models of the current era. We present a systematic analysis using modern techniques from the intersection of causality and AI/ML in a humble effort of reviving research around BPs. Specifically, we first compile the BPs into a Markov decision process, then secondly pose causal assumptions on the data generating process arguing for their applicability to BPs, and finally apply reinforcement learning techniques for solving the BPs subject to the causal assumptions.