Guiding Inferences in Connection Tableau by Recurrent Neural Networks
This work addresses automated theorem proving for researchers in logic and AI, but it appears incremental as it applies an existing method (RNNs) to a known bottleneck in proof guidance.
The authors tackled the problem of guiding clause selection in connection tableau proof calculus by using recurrent neural networks (RNNs) to encode literals and select clauses, achieving results compared to state-of-the-art gradient boosted trees.
We present a dataset and experiments on applying recurrent neural networks (RNNs) for guiding clause selection in the connection tableau proof calculus. The RNN encodes a sequence of literals from the current branch of the partial proof tree to a hidden vector state; using it, the system selects a clause for extending the proof tree. The training data and learning setup are described, and the results are discussed and compared with state of the art using gradient boosted trees. Additionally, we perform a conjecturing experiment in which the RNN does not just select an existing clause, but completely constructs the next tableau goal.