Approaches and Applications of Inductive Programming (Dagstuhl Seminar 23442)

Abstract

The Dagstuhl Seminar "Approaches and Applications of Inductive Programming" (AAIP) has taken place for the sixth time. The Dagstuhl Seminar series brings together researchers concerned with learning programs from input/output examples from different areas, mostly from machine learning and other branches of artificial intelligence research, cognitive scientists interested in human learning in complex domains, and researchers with a background in formal methods and programming languages. Main topics adressed in the AAIP 2023 seminar have been neurosymbolic approaches to IP bringing together learning and reasoning, IP as a post-hoc approach to explaining decision-making of deep learning blackbox models, and exploring the potential of deep learning approaches, especially large language models such as OpenAI Codex for IP. Topics discussed in working groups were Large Language Models and inductive programming in cognitive architectures, avoiding too much search in inductive programming, finding suitable benchmark problems, and evaluation criteria for interpretability and explainability of inductive programming.