WG5 Machine Learning in Proofs

The aim of this WG is to contribute to the field of machine-learning-based methods to improve the efficiency of automated theorem proving systems in terms of further development of techniques for proof guidance and premise selection. Furthemore, the group will explore how and to what extent tasks of computer-assisted reasoning can be extended to proofs that are represented in (controlled) natural languages.

Activities

• 6-8 September 2023: Joint WG4-WG5 meeting/Workshop on Natural Formal Mathematics, and on libraries of formal proofs and natural mathematical language, Cambridge, UK

• 18-20 April 2023: WG meeting co-organized with Joint Workshops (PAMLTP) and (DG4D^3), Prague (Czechia)

• 5-8 September 2022: WG meeting Machine Learning for Theorem Proving co-organized with AITP’22 conference, Aussois (France)