Abstract
Recent works on Large Language Models (LLMs) have demonstrated their effectiveness in learning general policies in automated planning. In particular, a system called PlanGPT has achieved impressive performance in terms of coverage in various domains. However, it may produce invalid plans that either satisfy only some goal fluents of the corresponding planning problem or violate the planned actions’ preconditions. To overcome this limitation, we propose a novel neuro-symbolic approach that combines PlanGPT with a planner capable of repairing (or completing) the plan generated by PlanGPT, thereby leveraging model-based reasoning. When PlanGPT generates a candidate plan for a specific planning problem, we validate it using a symbolic validator. If the generated plan is invalid, we execute the repair procedure of the planner LPG to obtain a valid solution plan from it. In this paper, we empirically evaluate the effectiveness of our approach and demonstrate its performances across various planning domains. Our results show significant improvements in the performance of both PlanGPT and LPG, highlighting the effectiveness of combining learning methods with traditional planning techniques.
M. Tummolo was enrolled in the National Doctorate on AI conducted by Sapienza, University of Rome with the University of Brescia.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Arora, D., Kambhampati, S.: Learning and leveraging verifiers to improve planning capabilities of pre-trained language models. arXiv preprint arXiv:2305.17077 (2023)
Chen, D., Thiébaux, S., Trevizan, F.: Learning domain-independent heuristics for grounded and lifted planning. In: Proceedings of 38th AAAI Conference on Artificial Intelligence (2024). http://felipe.trevizan.org/papers/chen24:goose.pdf
Chen, D., Trevizan, F., Thiébaux, S.: Return to tradition: learning reliable heuristics with classical machine learning. In: Proceedings of 34th International Conference on Automated Planning and Scheduling (ICAPS) (2024). http://felipe.trevizan.org/papers/chen24:wlkernel.pdf
Devlin, J., Chang, M., Lee, K., Toutanova, K.: BERT: pre-training of deep bidirectional transformers for language understanding. In: NAACL-HLT, pp. 4171–4186. Association for Computational Linguistics (2019)
Geib, C.W., Steedman, M.: On natural language processing and plan recognition. In: IJCAI, pp. 1612–1617 (2007)
Gerevini, A., Saetti, A., Serina, I.: Planning through stochastic local search and temporal action graphs in LPG. J. Artif. Intell. Res. 20, 239–290 (2003). https://doi.org/10.1613/jair.1183
Geva, M., Khashabi, D., Segal, E., Khot, T., Roth, D., Berant, J.: Did Aristotle use a laptop? A question answering benchmark with implicit reasoning strategies. Trans. Assoc. Comput. Linguistics 9, 346–361 (2021)
Hao, M., Trevizan, F., Thiébaux, S., Ferber, P., Hoffmann, J.: Guiding GBFS through learned pairwise rankings. In: Proceedings of 33rd International Joint Conference on AI (IJCAI) (2024). http://felipe.trevizan.org/papers/hao24:ranking.pdf
Helmert, M.: The fast downward planning system. J. Artif. Intell. Res. 26, 191–246 (2006)
Howey, R., Long, D., Fox, M.: VAL: automatic plan validation, continuous effects and mixed initiative planning using PDDL. In: ICTAI, pp. 294–301. IEEE Computer Society (2004)
Hu, Y., De Giacomo, G.: Generalized planning: synthesizing plans that work for multiple environments. In: IJCAI, pp. 918–923. IJCAI.org (2011)
Jiang, Z., Xu, F.F., Araki, J., Neubig, G.: How can we know what language models know. Trans. Assoc. Comput. Linguistics 8, 423–438 (2020)
Kautz, H., Selman, B.: Pushing the envelope: planning, propositional logic, and stochastic search. In: Proceedings of the National Conference on Artificial Intelligence, vol. 2 (1999)
Pallagani, V., et al.: Plansformer: generating symbolic plans using transformers. arXiv preprint arXiv:2212.08681 (2022)
Pallagani, V., et al.: Plansformer tool: demonstrating generation of symbolic plans using transformers. In: IJCAI, pp. 7158–7162. IJCAI.org (2023)
Petroni, F., et al.: Language models as knowledge bases? In: EMNLP/IJCNLP, pp. 2463–2473. Association for Computational Linguistics (2019)
Radford, A., Narasimhan, K.: Improving language understanding by generative pre-training (2018). https://api.semanticscholar.org/CorpusID:49313245
Radford, A., Wu, J., Child, R., Luan, D., Amodei, D., Sutskever, I.: Language models are unsupervised multitask learners (2019). https://api.semanticscholar.org/CorpusID:160025533
Richter, S., Westphal, M.: The LAMA planner: guiding cost-based anytime planning with landmarks. J. Artif. Intell. Res. 39, 127–177 (2010)
Rossetti, N., et al.: Learning general policies for planning through GPT models. In: 34th International Conference on Automated Planning and Scheduling (2024)
Rossetti, N., Tummolo, M., Gerevini, A.E., Putelli, L., Serina, I., Olivato, M.: Enhancing GPT-based planning policies by model-based plan validation. In: Proceedings of the 18th International Conference on Neural-Symbolic Learning and Reasoning (2024)
Seipp, J., Torralba, A., Hoffmann, J.: PDDL generators (2022). https://github.com/AI-Planning/pddl-generators
Serina, L., Chiari, M., Gerevini, A.E., Putelli, L., Serina, I.: A preliminary study on BERT applied to automated planning. In: Proceedings of the 10th Italian workshop on Planning and Scheduling (IPS 2022), RCRA Incontri E Confronti (RiCeRcA 2022), and the workshop on Strategies, Prediction, Interaction, and Reasoning in Italy (SPIRIT 2022) co-located with 21st International Conference of the Italian Association for Artificial Intelligence (AI*IA 2022), November 28 – December 2, 2022, University of Udine, Udine, Italy. CEUR Workshop Proceedings, vol. 3345. CEUR-WS.org (2022)
Serina, L., Putelli, L., Gerevini, A.E., Serina, I.: Synonyms, antonyms and factual knowledge in BERT heads. Future Internet 15(7), 230 (2023). https://doi.org/10.3390/fi15070230
Srivastava, S., Immerman, N., Zilberstein, S.: Learning generalized plans using abstract counting. In: AAAI, pp. 991–997. AAAI Press (2008)
Srivastava, S., Immerman, N., Zilberstein, S.: A new representation and associated algorithms for generalized planning. Artif. Intell. 175(2), 615–647 (2011)
Ståhlberg, S., Bonet, B., Geffner, H.: Learning general optimal policies with graph neural networks: expressive power, transparency, and limits. In: ICAPS, pp. 629–637. AAAI Press (2022)
Ståhlberg, S., Bonet, B., Geffner, H.: Learning generalized policies without supervision using GNNs. In: KR, pp. 474–483. IJCAI Org. (2022)
Stahlberg, S., Bonet, B., Geffner, H.: Muninn (2023). https://ipc2023-learning.github.io/abstracts/muninn.pdf
Valmeekam, K., Hernandez, A.O., Sreedharan, S., Kambhampati, S.: Large language models still can’t plan, a benchmark for LLMs on planning and reasoning about change. arXiv preprint arXiv:2206.10498 (2022)
Valmeekam, K., Marquez, M., Sreedharan, S., Kambhampati, S.: On the planning abilities of large language models - a critical investigation. In: NeurIPS (2023)
Valmeekam, K., Sreedharan, S., Marquez, M., Hernandez, A.O., Kambhampati, S.: On the planning abilities of large language models, a critical investigation with a proposed benchmark. arXiv preprint arXiv:2302.06706 (2023)
Vaswani, A., et al.: Attention is all you need. In: NIPS, pp. 5998–6008. Curran Associates Inc. (2017)
Wang, Y., Wang, W., Joty, S.R., Hoi, S.C.H.: CodeT5: identifier-aware unified pre-trained encoder-decoder models for code understanding and generation. In: EMNLP, pp. 8696–8708. Association for Computational Linguistics (2021)
Acknowledgements
This work was been supported by EU H2020 project AIPlan4EU (GA 101016442), EU ICT-48 2020 project TAILOR (GA 952215), MUR PRIN project RIPER (No. 20203FFYLK), Climate Change AI project (No. IG-2023-174), Regione Lombardia through the initiatives “Il Piano Lombardia - Interventi per la ripresa economica” and “Programma degli interventi per la ripresa economica: sviluppo di nuovi accordi di collaborazione con le università per la ricerca, l’innovazione e il trasferimento tecnologico” - DGR n. XI/4445/2021 and by SERICS (PE00000014) under the MUR National Recovery and Resilience Plan funded by the European Union - NextGenerationEU.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2025 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Tummolo, M., Rossetti, N., Gerevini, A.E., Olivato, M., Putelli, L., Serina, I. (2025). Integrating Classical Planners with GPT-Based Planning Policies. In: Artale, A., Cortellessa, G., Montali, M. (eds) AIxIA 2024 – Advances in Artificial Intelligence. AIxIA 2024. Lecture Notes in Computer Science(), vol 15450. Springer, Cham. https://doi.org/10.1007/978-3-031-80607-0_24
Download citation
DOI: https://doi.org/10.1007/978-3-031-80607-0_24
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-80606-3
Online ISBN: 978-3-031-80607-0
eBook Packages: Computer ScienceComputer Science (R0)