MADE: Benchmark Environments for Closed-Loop Materials Discovery

Shreshth A Malik, Tiarnan Doherty, Panagiotis Tigas, Muhammed Razzak, Stephen J. Roberts, Aron Walsh, Yarin Gal
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:86002-86026, 2026.

Abstract

Existing benchmarks for computational materials discovery primarily evaluate static predictive tasks or isolated computational sub-tasks. While valuable, these evaluations neglect the inherently iterative and adaptive nature of scientific discovery. We introduce MAterials Discovery Environments (MADE), a novel framework for benchmarking end-to-end autonomous materials discovery pipelines. MADE simulates closed loop discovery campaigns in which an agent or algorithm proposes, evaluates, and refines candidate materials under a constrained oracle budget, capturing the sequential and resource-limited nature of real discovery workflows. We formalize discovery as a search for thermodynamically stable compounds relative to a given convex hull, and evaluate efficacy and efficiency via comparison to baseline algorithms. The framework is flexible; users can compose discovery agents from interchangeable components such as generative models, filters, and planners, enabling the study of arbitrary workflows ranging from fixed pipelines to agentic systems. We demonstrate this by conducting systematic experiments across a diverse range of systems and algorithms, finding that adaptive planning becomes more important to discovery efficiency as the search space scales.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-malik26a, title = {{MADE}: Benchmark Environments for Closed-Loop Materials Discovery}, author = {Malik, Shreshth A and Doherty, Tiarnan and Tigas, Panagiotis and Razzak, Muhammed and Roberts, Stephen J. and Walsh, Aron and Gal, Yarin}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {86002--86026}, 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/malik26a/malik26a.pdf}, url = {https://proceedings.mlr.press/v306/malik26a.html}, abstract = {Existing benchmarks for computational materials discovery primarily evaluate static predictive tasks or isolated computational sub-tasks. While valuable, these evaluations neglect the inherently iterative and adaptive nature of scientific discovery. We introduce MAterials Discovery Environments (MADE), a novel framework for benchmarking end-to-end autonomous materials discovery pipelines. MADE simulates closed loop discovery campaigns in which an agent or algorithm proposes, evaluates, and refines candidate materials under a constrained oracle budget, capturing the sequential and resource-limited nature of real discovery workflows. We formalize discovery as a search for thermodynamically stable compounds relative to a given convex hull, and evaluate efficacy and efficiency via comparison to baseline algorithms. The framework is flexible; users can compose discovery agents from interchangeable components such as generative models, filters, and planners, enabling the study of arbitrary workflows ranging from fixed pipelines to agentic systems. We demonstrate this by conducting systematic experiments across a diverse range of systems and algorithms, finding that adaptive planning becomes more important to discovery efficiency as the search space scales.} }
Endnote
%0 Conference Paper %T MADE: Benchmark Environments for Closed-Loop Materials Discovery %A Shreshth A Malik %A Tiarnan Doherty %A Panagiotis Tigas %A Muhammed Razzak %A Stephen J. Roberts %A Aron Walsh %A Yarin Gal %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-malik26a %I PMLR %P 86002--86026 %U https://proceedings.mlr.press/v306/malik26a.html %V 306 %X Existing benchmarks for computational materials discovery primarily evaluate static predictive tasks or isolated computational sub-tasks. While valuable, these evaluations neglect the inherently iterative and adaptive nature of scientific discovery. We introduce MAterials Discovery Environments (MADE), a novel framework for benchmarking end-to-end autonomous materials discovery pipelines. MADE simulates closed loop discovery campaigns in which an agent or algorithm proposes, evaluates, and refines candidate materials under a constrained oracle budget, capturing the sequential and resource-limited nature of real discovery workflows. We formalize discovery as a search for thermodynamically stable compounds relative to a given convex hull, and evaluate efficacy and efficiency via comparison to baseline algorithms. The framework is flexible; users can compose discovery agents from interchangeable components such as generative models, filters, and planners, enabling the study of arbitrary workflows ranging from fixed pipelines to agentic systems. We demonstrate this by conducting systematic experiments across a diverse range of systems and algorithms, finding that adaptive planning becomes more important to discovery efficiency as the search space scales.
APA
Malik, S.A., Doherty, T., Tigas, P., Razzak, M., Roberts, S.J., Walsh, A. & Gal, Y.. (2026). MADE: Benchmark Environments for Closed-Loop Materials Discovery. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:86002-86026 Available from https://proceedings.mlr.press/v306/malik26a.html.

Related Material