Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
97 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Integrating Explanations in Learning LTL Specifications from Demonstrations (2404.02872v1)

Published 3 Apr 2024 in cs.AI

Abstract: This paper investigates whether recent advances in LLMs can assist in translating human explanations into a format that can robustly support learning Linear Temporal Logic (LTL) from demonstrations. Both LLMs and optimization-based methods can extract LTL specifications from demonstrations; however, they have distinct limitations. LLMs can quickly generate solutions and incorporate human explanations, but their lack of consistency and reliability hampers their applicability in safety-critical domains. On the other hand, optimization-based methods do provide formal guarantees but cannot process natural language explanations and face scalability challenges. We present a principled approach to combining LLMs and optimization-based methods to faithfully translate human explanations and demonstrations into LTL specifications. We have implemented a tool called Janaka based on our approach. Our experiments demonstrate the effectiveness of combining explanations with demonstrations in learning LTL specifications through several case studies.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (50)
  1. E. Abadi and R. I. Brafman, “Learning and solving regular decision processes,” in Proceedings of the Twenty-Ninth International Joint Conference on Artificial Intelligence, IJCAI.   ijcai.org, 2020, pp. 1948–1954. [Online]. Available: https://doi.org/10.24963/ijcai.2020/270
  2. T. Achterberg, “What’s new in gurobi 9.0,” Webinar Talk url: https://www. gurobi. com/wp-content/uploads/2019/12/Gurobi-90-Overview-Webinar-Slides-1. pdf, 2019.
  3. M. Afzal, S. Gambhir, A. Gupta, S. Krishna, A. Trivedi, and A. Velasquez, “Ltl-based non-markovian inverse reinforcement learning,” in Proceedings of the 2023 International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2023, London, United Kingdom, 29 May 2023 - 2 June 2023, N. Agmon, B. An, A. Ricci, and W. Yeoh, Eds.   ACM, 2023, pp. 2857–2859. [Online]. Available: https://dl.acm.org/doi/10.5555/3545946.3599102
  4. R. Alur, O. Bastani, K. Jothimurugan, M. Perez, F. Somenzi, and A. Trivedi, “Policy synthesis and reinforcement learning for discounted ltl,” arXiv preprint arXiv:2305.17115, 2023.
  5. D. Angluin, “Learning regular sets from queries and counterexamples,” Information and computation, vol. 75, no. 2, pp. 87–106, 1987.
  6. H. Bensusan, “God doesn’t always shave with occam’s razor—learning when and how to prune,” in Machine Learning: ECML-98: 10th European Conference on Machine Learning Chemnitz, Germany, April 21–23, 1998 Proceedings 10.   Springer, 1998, pp. 119–124.
  7. A. Blumer, A. Ehrenfeucht, D. Haussler, and M. K. Warmuth, “Occam’s razor,” Information processing letters, vol. 24, no. 6, pp. 377–380, 1987.
  8. A. Bohy, V. Bruyère, E. Filiot, N. Jin, and J.-F. Raskin, “Acacia+, a tool for ltl synthesis,” in Computer Aided Verification: 24th International Conference, CAV 2012, Berkeley, CA, USA, July 7-13, 2012 Proceedings 24.   Springer, 2012, pp. 652–657.
  9. G. Brockman, V. Cheung, L. Pettersson, J. Schneider, J. Schulman, J. Tang, and W. Zaremba, “Openai gym,” 2016.
  10. A. Brunello, A. Montanari, and M. Reynolds, “Synthesis of ltl formulas from natural language texts: State of the art and research directions,” in 26th International symposium on temporal representation and reasoning (TIME 2019).   Schloss Dagstuhl-Leibniz-Zentrum fuer Informatik, 2019.
  11. A. Camacho, R. T. Icarte, T. Q. Klassen, R. A. Valenzano, and S. A. McIlraith, “Ltl and beyond: Formal languages for reward function specification in reinforcement learning.” in IJCAI, vol. 19, 2019, pp. 6065–6073.
  12. R. C. Carrasco and J. Oncina, “Learning stochastic regular grammars by means of a state merging method,” in International Colloquium on Grammatical Inference.   Springer, 1994, pp. 139–152.
  13. G. Carvalho, F. Barros, A. Carvalho, A. Cavalcanti, A. Mota, and A. Sampaio, “Nat2test tool: From natural language requirements to test cases based on csp,” in Software Engineering and Formal Methods: 13th International Conference, SEFM 2015, York, UK, September 7-11, 2015. Proceedings.   Springer, 2015, pp. 283–290.
  14. M. Cosler, C. Hahn, D. Mendoza, F. Schmitt, and C. Trippel, “nl2spec: Interactively translating unstructured natural language to temporal logics with large language models,” arXiv preprint arXiv:2303.04864, 2023.
  15. G. De Giacomo and M. Vardi, “Synthesis for ltl and ldl on finite traces,” in Twenty-Fourth International Joint Conference on Artificial Intelligence, 2015.
  16. G. De Giacomo and M. Y. Vardi, “Linear temporal logic and linear dynamic logic on finite traces,” in IJCAI’13 Proceedings of the Twenty-Third international joint conference on Artificial Intelligence.   Association for Computing Machinery, 2013, pp. 854–860.
  17. P. Domingos, “The role of occam’s razor in knowledge discovery,” Data mining and knowledge discovery, vol. 3, pp. 409–425, 1999.
  18. A. Duret-Lutz, A. Lewkowicz, A. Fauchille, T. Michaud, E. Renault, and L. Xu, “Spot 2.0—a framework for ltl and-automata manipulation,” in International Symposium on Automated Technology for Verification and Analysis.   Springer, 2016, pp. 122–129.
  19. S. Ghosh, D. Elenius, W. Li, P. Lincoln, N. Shankar, and W. Steiner, “Arsenal: automatic requirements specification extraction from natural language,” in NASA Formal Methods: 8th International Symposium, NFM 2016, Minneapolis, MN, USA, June 7-9, 2016, Proceedings 8.   Springer, 2016, pp. 41–46.
  20. D. Giannakopoulou, T. Pressburger, A. Mavridou, and J. Schumann, “Generation of formal requirements from structured natural language,” in Requirements Engineering: Foundation for Software Quality: 26th International Working Conference, REFSQ 2020, Pisa, Italy, March 24–27, 2020, Proceedings 26.   Springer, 2020, pp. 19–35.
  21. E. M. Gold, “Language identification in the limit,” Information and control, vol. 10, no. 5, pp. 447–474, 1967.
  22. C. Hahn, F. Schmitt, J. J. Tillman, N. Metzger, J. Siber, and B. Finkbeiner, “Formal specifications from natural language,” CoRR, vol. abs/2206.01962, 2022. [Online]. Available: https://doi.org/10.48550/arXiv.2206.01962
  23. E. M. Hahn, M. Perez, S. Schewe, F. Somenzi, A. Trivedi, and D. Wojtczak, “Omega-regular objectives in model-free reinforcement learning,” in International conference on tools and algorithms for the construction and analysis of systems.   Springer, 2019, pp. 395–412.
  24. J. J. Horning, “A study of grammatical inference.” Stanford University California, Department of Computer Science, Tech. Rep., 1969.
  25. M. Ilieva and O. Ormandjieva, “Automatic transition of natural language software requirements specification into formal presentation,” in International Conference on Application of Natural Language to Information Systems.   Springer, 2005, pp. 392–397.
  26. J. Kim, C. Muise, A. J. Shah, S. Agarwal, and J. A. Shah, “Bayesian inference of linear temporal logic specifications for contrastive explanations.”   International Joint Conferences on Artificial Intelligence, 2019.
  27. J. Křetínskỳ, T. Meggendorfer, and S. Sickert, “Owl: a library for-words, automata, and ltl,” in International Symposium on Automated Technology for Verification and Analysis.   Springer, 2018, pp. 543–550.
  28. C. Lemieux, D. Park, and I. Beschastnikh, “General ltl specification mining (t),” in 2015 30th IEEE/ACM International Conference on Automated Software Engineering (ASE).   IEEE, 2015, pp. 81–92.
  29. W. Li, L. Dworkin, and S. A. Seshia, “Mining assumptions for synthesis,” in Ninth ACM/IEEE International Conference on Formal Methods and Models for Codesign (MEMPCODE2011).   IEEE, 2011, pp. 43–50.
  30. J. X. Liu, Z. Yang, I. Idrees, S. Liang, B. Schornstein, S. Tellex, and A. Shah, “Lang2ltl: Translating natural language commands to temporal robot task specification,” arXiv preprint arXiv:2302.11649, 2023.
  31. M. Lopes, F. Melo, and L. Montesano, “Active learning for reward estimation in inverse reinforcement learning,” in Joint European Conference on Machine Learning and Knowledge Discovery in Databases.   Springer, 2009, pp. 31–46.
  32. H. Mao, Y. Chen, M. Jaeger, T. D. Nielsen, K. G. Larsen, and B. Nielsen, “Learning deterministic probabilistic automata from a model checking perspective,” Machine Learning, vol. 105, no. 2, pp. 255–299, 2016.
  33. ——, “Learning markov decision processes for model checking,” Electronic Proceedings in Theoretical Computer Science, vol. 103, pp. 49–63, 2012.
  34. A. Mavrogiannis, C. Mavrogiannis, and Y. Aloimonos, “Cook2ltl: Translating cooking recipes to ltl formulae using large language models,” arXiv preprint arXiv:2310.00163, 2023.
  35. D. Neider and I. Gavran, “Learning linear temporal properties,” in 2018 Formal Methods in Computer Aided Design (FMCAD), 2018, pp. 1–10.
  36. R. OpenAI, “Gpt-4 technical report,” arXiv, pp. 2303–08 774, 2023.
  37. S. E. Palmer, “Visual perception of objects,” Handbook of psychology, pp. 177–211, 2003.
  38. A. Radford, K. Narasimhan, T. Salimans, I. Sutskever et al., “Improving language understanding by generative pre-training,” 2018.
  39. K. Rahmani, M. Raza, S. Gulwani, V. Le, D. Morris, A. Radhakrishna, G. Soares, and A. Tiwari, “Multi-modal program inference: a marriage of pre-trained language models and component-based synthesis,” Proceedings of the ACM on Programming Languages, vol. 5, no. OOPSLA, pp. 1–29, 2021.
  40. R. Roy, J.-R. Gaglione, N. Baharisangari, D. Neider, Z. Xu, and U. Topcu, “Learning interpretable temporal properties from positive examples only,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 5, 2023, pp. 6507–6515.
  41. D. Sadigh, E. S. Kim, S. Coogan, S. S. Sastry, and S. A. Seshia, “A learning based approach to control synthesis of markov decision processes for linear temporal logic specifications,” in 53rd IEEE Conference on Decision and Control.   IEEE, 2014, pp. 1091–1096.
  42. D. Sadoun, C. Dubois, Y. Ghamri-Doudane, and B. Grau, “From natural language requirements to formal specification using an ontology,” in 2013 IEEE 25th International Conference on Tools with Artificial Intelligence.   IEEE, 2013, pp. 755–760.
  43. T. Santos, G. Carvalho, and A. Sampaio, “Formal modelling of environment restrictions from natural-language requirements,” in Formal Methods: Foundations and Applications: 21st Brazilian Symposium, SBMF 2018, Salvador, Brazil, November 26–30, 2018, Proceedings 21.   Springer, 2018, pp. 252–270.
  44. P. Tabuada and D. Neider, “Robust linear temporal logic,” arXiv preprint arXiv:1510.08970, 2015.
  45. A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. u. Kaiser, and I. Polosukhin, “Attention is all you need,” in Advances in Neural Information Processing Systems, I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, Eds., vol. 30.   Curran Associates, Inc., 2017. [Online]. Available: https://proceedings.neurips.cc/paper_files/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf
  46. M. Vazquez-Chanlatte, S. Jha, A. Tiwari, M. K. Ho, and S. Seshia, “Learning task specifications from demonstrations,” Advances in neural information processing systems, vol. 31, 2018.
  47. J. Wagemans, J. H. Elder, M. Kubovy, S. E. Palmer, M. A. Peterson, M. Singh, and R. Von der Heydt, “A century of gestalt psychology in visual perception: I. perceptual grouping and figure–ground organization.” Psychological bulletin, vol. 138, no. 6, p. 1172, 2012.
  48. M. W. Whalen, A. Gacek, D. Cofer, A. Murugesan, M. P. Heimdahl, and S. Rayadurgam, “Your” what” is my” how”: Iteration and hierarchy in system design,” IEEE software, vol. 30, no. 2, pp. 54–60, 2012.
  49. R. Yan, C.-H. Cheng, and Y. Chai, “Formal consistency checking over specifications in natural languages,” in 2015 Design, Automation & Test in Europe Conference & Exhibition (DATE).   IEEE, 2015, pp. 1677–1682.
  50. Y. Zhang, Y. Li, L. Cui, D. Cai, L. Liu, T. Fu, X. Huang, E. Zhao, Y. Zhang, Y. Chen et al., “Siren’s song in the ai ocean: A survey on hallucination in large language models,” arXiv preprint arXiv:2309.01219, 2023.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (6)
  1. Ashutosh Gupta (27 papers)
  2. John Komp (3 papers)
  3. Abhay Singh Rajput (1 paper)
  4. Krishna Shankaranarayanan (1 paper)
  5. Ashutosh Trivedi (76 papers)
  6. Namrita Varshney (1 paper)

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now