ENIGMA-NG: Efficient Neural and Gradient-Boosted Inference Guidance for E
This work addresses efficiency challenges in automated theorem proving for AI/formal verification, representing an incremental improvement over previous ENIGMA methods.
The paper tackles the problem of clause guidance in automated theorem provers by extending the ENIGMA approach with gradient-boosted trees and recursive neural networks, showing that deep integration with ATP data-structures leads to competitive real-time results and improves on manually designed guidance.
We describe an efficient implementation of clause guidance in saturation-based automated theorem provers extending the ENIGMA approach. Unlike in the first ENIGMA implementation where fast linear classifier is trained and used together with manually engineered features, we have started to experiment with more sophisticated state-of-the-art machine learning methods such as gradient boosted trees and recursive neural networks. In particular the latter approach poses challenges in terms of efficiency of clause evaluation, however, we show that deep integration of the neural evaluation with the ATP data-structures can largely amortize this cost and lead to competitive real-time results. Both methods are evaluated on a large dataset of theorem proving problems and compared with the previous approaches. The resulting methods improve on the manually designed clause guidance, providing the first practically convincing application of gradient-boosted and neural clause guidance in saturation-style automated theorem provers.