Polymer property prediction improved with new open benchmark dataset
An open benchmark for machine learning-based polymer property prediction
Machine LearningArtificial Intelligence
Summary
Predicting how polymers behave is hard because existing tools only focus on simple polymers and there is no big, open dataset to test new methods. The authors created a large public dataset called PolyBench26 with about 250,000 data points that cover many types of polymers and physical properties. They tested different machine learning approaches on this dataset and found that graph-based models predict properties more accurately and handle complex polymer structures better. This new benchmark helps improve and compare models for designing a wider variety of polymers.
What this means in practice
- •For materials engineers: Use the PolyBench26 dataset to train machine learning models that predict polymer properties for faster materials design.
- •For chemical manufacturing teams: Develop quality control processes that leverage improved polymer property predictions across diverse polymer types using graph-based models.
Authors
Robert W. Learsch, Nicholas Liesen, Daniel S. Levine, Anna M. Hiszpanski, Evan R. Antoniuk
Abstract
Polymer property prediction lacks open, standardized benchmarks that enable rigorous comparison of machine-learning methods, with existing resources covering only a narrow fraction of polymer architectures, such as homopolymers. We introduce Polymer Benchmark 2026 (PolyBench26), an open dataset comprising nearly 250,000 polymer-property datapoints across eight physical properties, including data from experimental measurements, density functional theory, and molecular dynamics. The benchmark supports four evaluation tasks across homopolymers and alternating, random, and block copolymers: in-distribution property prediction, dataset-size scaling, repeat-unit complexity, and transfer to held-out polymer architectures. We compare language model, graph-based, and descriptor-based approaches and find graph-based models provide the lowest errors in property prediction, retain their advantage across the evaluated training-set sizes, and remain robust to increasing repeat-unit complexity. PolyBench26 provides a reproducible foundation for developing models for the increasingly complex polymer design space. The PolyBench26 benchmark is available open-source at https://github.com/rlearsch/PolymerBenchmark2026.