Automated Proof Synthesis for the Minimal Propositional Logic with Deep Neural Networks

  • Taro SekiyamaEmail author
  • Kohei Suenaga
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11275)


This work explores the application of deep learning, a machine learning technique that uses deep neural networks (DNN) in its core, to an automated theorem proving (ATP) problem. To this end, we construct a statistical model which quantifies the likelihood that a proof is indeed a correct one of a given proposition. Based on this model, we give a proof-synthesis procedure that searches for a proof in the order of the likelihood. This procedure uses an estimator of the likelihood of an inference rule being applied at each step of a proof. As an implementation of the estimator, we propose a proposition-to-proof architecture, which is a DNN tailored to the automated proof synthesis problem. To empirically demonstrate its usefulness, we apply our model to synthesize proofs of the minimal propositional logic. We train the proposition-to-proof model using a training dataset of proposition–proof pairs. The evaluation against a benchmark set shows the very high accuracy and an improvement to the recent work of neural proof synthesis.


Automated theorem proving Deep learning Neural networks 



We are grateful to Akifumi Imanishi, Kensuke Kojima, and Takayuki Muranushi for the discussion and the contribution to the earlier version of the present work. Kohei Suenaga is supported by JST PRESTO Grant Number JPMJPR15E5, Japan. Taro Sekiyama is supported by ERATO HASUO Metamathematics for Systems Design Project (No. JPMJER1603), JST.


  1. 1.
    Avellone, A., Fiorino, G., Moscato, U.: A new \(O(n \log {n})\)-space decision procedure for propositional intuitionistic logic. Kurt Gödel Society Collegium Logicum VIII, 17–33 (2004)Google Scholar
  2. 2.
    Bahdanau, D., Cho, K., Bengio, Y.: Neural machine translation by jointly learning to align and translate. CoRR abs/1409.0473 (2014)Google Scholar
  3. 3.
    Barnett, M., et al.: The Spec# programming system: challenges and directions. In: Meyer, B., Woodcock, J. (eds.) VSTTE 2005. LNCS, vol. 4171, pp. 144–152. Springer, Heidelberg (2008). Scholar
  4. 4.
    Ben-Yelles, C.: Type assignment in the lambda-calculus: syntax and semantics. Ph.D. thesis, University College of Swansea, September 1979Google Scholar
  5. 5.
    Bibel, W.: Automated Theorem Proving. Springer, Heidelberg (2013)Google Scholar
  6. 6.
    Bünz, B., Lamm, M.: Graph neural networks and boolean satisfiability. CoRR abs/1702.03592 (2017)Google Scholar
  7. 7.
    Chalin, P., James, P.R., Karabotsos, G.: An integrated verification environment for JML: architecture and early results. In: Proceedings of SAVCBS, pp. 47–53. ACM, New York (2007)Google Scholar
  8. 8.
    Evans, R., Saxton, D., Amos, D., Kohli, P., Grefenstette, E.: Can neural networks understand logical entailment? CoRR abs/1802.08535 (2018)Google Scholar
  9. 9.
    Filliâtre, J.-C., Paskevich, A.: Why3 — where programs meet provers. In: Felleisen, M., Gardner, P. (eds.) ESOP 2013. LNCS, vol. 7792, pp. 125–128. Springer, Heidelberg (2013). Scholar
  10. 10.
    Fukushima, K.: Neocognitron: a self-organizing neural network for a mechanism of pattern recognition unaffected by shift in position. Biol. Cybern. 36, 193–202 (1980)CrossRefGoogle Scholar
  11. 11.
    Glorot, X., Bordes, A., Bengio, Y.: Deep sparse rectifier neural networks. In: Proceedings of AISTATS, pp. 315–323 (2011)Google Scholar
  12. 12.
    Goodfellow, I., Bengio, Y., Courville, A.: Deep Learning. MIT Press, Cambridge (2016)zbMATHGoogle Scholar
  13. 13.
    Harrison, J.: HOL light: an overview. In: Berghofer, S., Nipkow, T., Urban, C., Wenzel, M. (eds.) TPHOLs 2009. LNCS, vol. 5674, pp. 60–66. Springer, Heidelberg (2009). Scholar
  14. 14.
    He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: Proceedings of CVPR, pp. 770–778 (2016)Google Scholar
  15. 15.
    Hinton, G., et al.: Deep neural networks for acoustic modeling in speech recognition: the shared views of four research groups. IEEE Sig. Process. Mag. 29(6), 82–97 (2012)CrossRefGoogle Scholar
  16. 16.
    Irving, G., Szegedy, C., Alemi, A.A., Eén, N., Chollet, F., Urban, J.: DeepMath - deep sequence models for premise selection. In: Proceedings of NIPS, pp. 2235–2243 (2016)Google Scholar
  17. 17.
    Johnson, J.L.: A neural network approach to the 3-satisfiability problem. J. Parallel Distrib. Comput. 6(2), 435–449 (1989)CrossRefGoogle Scholar
  18. 18.
    Kaliszyk, C., Chollet, F., Szegedy, C.: HolStep: a machine learning dataset for higher-order logic theorem proving. CoRR abs/1703.00426 (2017)Google Scholar
  19. 19.
    Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. CoRR abs/1412.6980 (2014)Google Scholar
  20. 20.
    Klein, G., et al.: seL4: formal verification of an OS kernel. In: Proceedings of SOSP, pp. 207–220 (2009)Google Scholar
  21. 21.
    Krizhevsky, A., Sutskever, I., Hinton, G.E.: Imagenet classification with deep convolutional neural networks. In: Proceedings of NIPS, pp. 1106–1114 (2012)Google Scholar
  22. 22.
    Leroy, X.: Formal verification of a realistic compiler. Commun. ACM 52(7), 107–115 (2009)CrossRefGoogle Scholar
  23. 23.
    Loos, S.M., Irving, G., Szegedy, C., Kaliszyk, C.: Deep network guided proof search. In: Proceedings of LPAR, pp. 85–105 (2017)Google Scholar
  24. 24.
    Milner, R.: A theory of type polymorphism in programming. J. Comput. Syst. Sci. 17(3), 348–375 (1978)MathSciNetCrossRefGoogle Scholar
  25. 25.
    Mou, L., Li, G., Zhang, L., Wang, T., Jin, Z.: Convolutional neural networks over tree structures for programming language processing. In: Proceedings of AAAI, pp. 1287–1293 (2016)Google Scholar
  26. 26.
    Raths, T., Otten, J., Kreitz, C.: The ILTP problem library for intuitionistic logic. J. Autom. Reason. 38(1–3), 261–271 (2007)MathSciNetCrossRefGoogle Scholar
  27. 27.
    Scarselli, F., Gori, M., Tsoi, A.C., Hagenbuchner, M., Monfardini, G.: The graph neural network model. IEEE Trans. Neural Netw. 20(1), 61–80 (2009)CrossRefGoogle Scholar
  28. 28.
    Schulz, S.: System description: E 1.8. In: McMillan, K., Middeldorp, A., Voronkov, A. (eds.) LPAR 2013. LNCS, vol. 8312, pp. 735–743. Springer, Heidelberg (2013). Scholar
  29. 29.
    Sekiyama, T., Imanishi, A., Suenaga, K.: Towards proof synthesis guided by neural machine translation for intuitionistic propositional logic. CoRR abs/1706.06462 (2017)Google Scholar
  30. 30.
    Sekiyama, T., Suenaga, K.: Automated proof synthesis for propositional logic with deep neural networks. CoRR abs/1805.11799 (2018).
  31. 31.
    Selsam, D., Lamm, M., Bünz, B., Liang, P., de Moura, L., Dill, D.L.: Learning a SAT solver from single-bit supervision. CoRR abs/1802.03685 (2018)Google Scholar
  32. 32.
    Sørensen, M.H., Urzyczyn, P.: Lectures on the Curry-Howard Isomorphism. Studies in Logic and the Foundations of Mathematics, vol. 149. Elsevier Science Inc., Amsterdam (2006)CrossRefGoogle Scholar
  33. 33.
    Sutskever, I., Vinyals, O., Le, Q.V.: Sequence to sequence learning with neural networks. In: Proceedings of NIPS, pp. 3104–3112 (2014)Google Scholar
  34. 34.
    Tokui, S., Oono, K., Hido, S., Clayton, J.: Chainer: a next-generation open source framework for deep learning. In: Proceedings of workshop on LearningSys (2015)Google Scholar
  35. 35.
    Wang, M., Tang, Y., Wang, J., Deng, J.: Premise selection for theorem proving by deep graph embedding. In: Proceedings of NIPS, pp. 2783–2793 (2017)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  1. 1.National Institute of InformaticsTokyoJapan
  2. 2.Kyoto UniversityKyotoJapan
  3. 3.JST PRESTOTokyoJapan

Personalised recommendations