Labeled data for imitation learning of theorem proving in large libraries of formalized
mathematics is scarce as such libraries require years of concentrated effort by human …

We introduce a theorem proving algorithm that uses practically no domain heuristics for
guiding its connection-style proof search. Instead, it runs many Monte-Carlo simulations …

Deep learning techniques lie at the heart of several significant AI advances in recent years
including object recognition and detection, image captioning, machine translation, speech …

We propose a deep learning-based approach to the problem of premise selection: selecting
mathematical statements relevant for proving a given conjecture. We represent a higher …

Automated theorem provers and formal proof assistants are general reasoning systems that
are in theory capable of proving arbitrarily hard theorems, thus solving arbitrary problems …

We describe an implementation of gradient boosting and neural guidance of saturation-style
automated theorem provers that does not depend on consistent symbol names across …

We describe an efficient implementation of given clause selection in saturation-based
automated theorem provers, extending the previous ENIGMA approach. Unlike in the first …

One of the main success stories of automatic theorem provers has been their integration into
proof assistants. Such integrations, or" hammers," increase proof automation and hence …

Automated reasoning and theorem proving have recently become major challenges for
machine learning. In other domains, representations that are able to abstract over …

We present a reinforcement learning toolkit for experiments with guiding automated theorem
proving in the connection calculus. The core of the toolkit is a compact and easy to extend …