Efficient Test-time Inference for Generative Planning Models with OCL Search

Robert Gieselmann, Mihai Samson, Federico Pecora, Jeremy L Wyatt
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:35065-35099, 2026.

Abstract

Generative models have emerged as a powerful paradigm for AI planning, yet their performance remains constrained by the training data distribution. One approach is to improve generated solutions during inference by scaling test-time compute. A more efficient alternative is to optimize the inference process itself. In this paper, we show that a modified version of a classical Open-Closed List (OCL) search provides just such an efficient inference procedure. Our algorithm synergizes two learned components: a generative model that performs fast rollouts from intermediate states and a heuristic model that prioritizes among candidate reasoning paths. Key contributions include novel exploration control mechanisms and integration of learned models within the OCL framework. Across multiple combinatorial planning domains, our approach outperforms both neurosymbolic search baselines and classical solvers in computational efficiency and solution quality.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-gieselmann26a, title = {Efficient Test-time Inference for Generative Planning Models with {OCL} Search}, author = {Gieselmann, Robert and Samson, Mihai and Pecora, Federico and Wyatt, Jeremy L}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {35065--35099}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/gieselmann26a/gieselmann26a.pdf}, url = {https://proceedings.mlr.press/v306/gieselmann26a.html}, abstract = {Generative models have emerged as a powerful paradigm for AI planning, yet their performance remains constrained by the training data distribution. One approach is to improve generated solutions during inference by scaling test-time compute. A more efficient alternative is to optimize the inference process itself. In this paper, we show that a modified version of a classical Open-Closed List (OCL) search provides just such an efficient inference procedure. Our algorithm synergizes two learned components: a generative model that performs fast rollouts from intermediate states and a heuristic model that prioritizes among candidate reasoning paths. Key contributions include novel exploration control mechanisms and integration of learned models within the OCL framework. Across multiple combinatorial planning domains, our approach outperforms both neurosymbolic search baselines and classical solvers in computational efficiency and solution quality.} }
Endnote
%0 Conference Paper %T Efficient Test-time Inference for Generative Planning Models with OCL Search %A Robert Gieselmann %A Mihai Samson %A Federico Pecora %A Jeremy L Wyatt %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-gieselmann26a %I PMLR %P 35065--35099 %U https://proceedings.mlr.press/v306/gieselmann26a.html %V 306 %X Generative models have emerged as a powerful paradigm for AI planning, yet their performance remains constrained by the training data distribution. One approach is to improve generated solutions during inference by scaling test-time compute. A more efficient alternative is to optimize the inference process itself. In this paper, we show that a modified version of a classical Open-Closed List (OCL) search provides just such an efficient inference procedure. Our algorithm synergizes two learned components: a generative model that performs fast rollouts from intermediate states and a heuristic model that prioritizes among candidate reasoning paths. Key contributions include novel exploration control mechanisms and integration of learned models within the OCL framework. Across multiple combinatorial planning domains, our approach outperforms both neurosymbolic search baselines and classical solvers in computational efficiency and solution quality.
APA
Gieselmann, R., Samson, M., Pecora, F. & Wyatt, J.L.. (2026). Efficient Test-time Inference for Generative Planning Models with OCL Search. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:35065-35099 Available from https://proceedings.mlr.press/v306/gieselmann26a.html.

Related Material