Improving ENIGMA-Style Clause Selection While Learning From History
This work addresses efficiency in automated theorem proving, which is incremental as it builds on existing ENIGMA methods.
The paper tackled the problem of clause selection in saturation-based theorem provers by improving the ENIGMA-style machine-learned guidance, resulting in a 41% improvement on a relevant subset of SMT-LIB in real-time evaluation.
We re-examine the topic of machine-learned clause selection guidance in saturation-based theorem provers. The central idea, recently popularized by the ENIGMA system, is to learn a classifier for recognizing clauses that appeared in previously discovered proofs. In subsequent runs, clauses classified positively are prioritized for selection. We propose several improvements to this approach and experimentally confirm their viability. For the demonstration, we use a recursive neural network to classify clauses based on their derivation history and the presence or absence of automatically supplied theory axioms therein. The automatic theorem prover Vampire guided by the network achieves a 41% improvement on a relevant subset of SMT-LIB in a real time evaluation.