Fig. 4: Generation of ROP using regression transformers.

a Flowchart depicting training process for the regression transformer (RT). The RT can predict physical and experimental properties of monomer-catalyst pairs (blue stream) or conditionally generate catalysts given a monomer and desired properties (yellow stream). The SMILES input and output of the RT are shown here, however the RT internally uses SELFIES representations⁷¹. See the Methods section for details. b Prediction performance for conversion, dispersity, and M_{n, GPC} properties of monomer–catalyst pairs from the test data set (blue circles). Solid blue line is the linear regression fit, shaded blue area represents 95% confidence for the linear regression fit, and dashed grey line is hypothetical perfect fit. All Pearson correlations were statistically significant (p < 0.001; two-sided; normality assumption). R² values are 0.66, 0.64 and 0.35 for conversion, dispersity and M_{n, GPC} respectively. The mean-absolute-errors are 0.12 (conversion), 0.10 (dispersity) and 0.26 (M_{n, GPC}). Note that M_{n, GPC} has been modeled on a log₁₀ scale. Conversion values are percentages plotted between 0 and 1, where 0 equals 0% conversion and 1 equals 100% conversion. c Tree manifold approximation and projection (TMAP) visualization⁷⁴ of generated catalysts and their physical properties (colored here by synthesizability scores, SAS). Source data for b and c are provided as a Source Data file.