Timezone: »

 
Poster
Reinforcement Learning of Theorem Proving
Cezary Kaliszyk · Josef Urban · Henryk Michalewski · Miroslav Olšák

Wed Dec 05 02:00 PM -- 04:00 PM (PST) @ Room 517 AB #110
in Wed Poster Session B »

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 guided by reinforcement learning from previous proof attempts. We produce several versions of the prover, parameterized by different learning and guiding algorithms. The strongest version of the system is trained on a large corpus of mathematical problems and evaluated on previously unseen problems. The trained system solves within the same number of inferences over 40% more problems than a baseline prover, which is an unusually high improvement in this hard AI domain. To our knowledge this is the first time reinforcement learning has been convincingly applied to solving general mathematical problems on a large scale.

Keywords: Reinforcement Learning and Planning · Reinforcement Learning · Problem Solving · Reasoning
[ Paper] 

Author Information

Cezary Kaliszyk (Innsbruck University)
Josef Urban (Czech Technical University in Prague)
Henryk Michalewski (University of Warsaw, University of Oxford, deepsense.ai)
Miroslav Olšák (Charles University in Prague)

More from the Same Authors