MADIL: An MDL-based Framework for Efficient Program Synthesis in the ARC Benchmark
It addresses the problem of high computational costs and lack of interpretability in AI systems for researchers and practitioners, though it is incremental as it builds on existing MDL principles for a specific benchmark.
The paper tackles efficient skill acquisition and generalization in AI by introducing MADIL, a Minimum Description Length-based framework for program synthesis on the ARC benchmark, achieving 7% performance at ArcPrize 2024 with improved efficiency and interpretability compared to LLM-based methods.
Artificial Intelligence (AI) has achieved remarkable success in specialized tasks but struggles with efficient skill acquisition and generalization. The Abstraction and Reasoning Corpus (ARC) benchmark evaluates intelligence based on minimal training requirements. While Large Language Models (LLMs) have recently improved ARC performance, they rely on extensive pre-training and high computational costs. We introduce MADIL (MDL-based AI), a novel approach leveraging the Minimum Description Length (MDL) principle for efficient inductive learning. MADIL performs pattern-based decomposition, enabling structured generalization. While its performance (7% at ArcPrize 2024) remains below LLM-based methods, it offers greater efficiency and interpretability. This paper details MADIL's methodology, its application to ARC, and experimental evaluations.