Learning Universal Predictors
This work addresses the fundamental problem of achieving universal prediction in AI, which could impact all of ML/AI by advancing meta-learning towards theoretical limits, though it is incremental in applying existing methods to new data.
The paper explores using meta-learning to train neural networks to approximate Solomonoff Induction, a universal predictor, by generating training data with Universal Turing Machines. The results indicate that UTM data is effective for meta-learning, enabling networks to learn universal prediction strategies across various architectures and data generators.
Meta-learning has emerged as a powerful approach to train neural networks to learn new tasks quickly from limited data. Broad exposure to different tasks leads to versatile representations enabling general problem solving. But, what are the limits of meta-learning? In this work, we explore the potential of amortizing the most powerful universal predictor, namely Solomonoff Induction (SI), into neural networks via leveraging meta-learning to its limits. We use Universal Turing Machines (UTMs) to generate training data used to expose networks to a broad range of patterns. We provide theoretical analysis of the UTM data generation processes and meta-training protocols. We conduct comprehensive experiments with neural architectures (e.g. LSTMs, Transformers) and algorithmic data generators of varying complexity and universality. Our results suggest that UTM data is a valuable resource for meta-learning, and that it can be used to train neural networks capable of learning universal prediction strategies.