Towards Learning Abductive Reasoning using VSA Distributed Representations
This addresses the problem of improving interpretability and efficiency in abductive reasoning for AI systems, though it appears incremental as it builds on existing methods like Learn-VRF.
The paper tackles abstract reasoning tasks by introducing ARLC, a model that achieves state-of-the-art accuracy on the I-RAVEN dataset for Raven's progressive matrices, including in out-of-distribution tests, while using far fewer parameters than baselines like large language models.
We introduce the Abductive Rule Learner with Context-awareness (ARLC), a model that solves abstract reasoning tasks based on Learn-VRF. ARLC features a novel and more broadly applicable training objective for abductive reasoning, resulting in better interpretability and higher accuracy when solving Raven's progressive matrices (RPM). ARLC allows both programming domain knowledge and learning the rules underlying a data distribution. We evaluate ARLC on the I-RAVEN dataset, showcasing state-of-the-art accuracy across both in-distribution and out-of-distribution (unseen attribute-rule pairs) tests. ARLC surpasses neuro-symbolic and connectionist baselines, including large language models, despite having orders of magnitude fewer parameters. We show ARLC's robustness to post-programming training by incrementally learning from examples on top of programmed knowledge, which only improves its performance and does not result in catastrophic forgetting of the programmed solution. We validate ARLC's seamless transfer learning from a 2x2 RPM constellation to unseen constellations. Our code is available at https://github.com/IBM/abductive-rule-learner-with-context-awareness.