Progress in science depends on new techniques, new discoveries and new ideas, probably in that order. - Sydney Brenner

we believe that building the right tools is essential for making progress in science. a focus of our work has been on developing benchmarks (“benches”) to understand the capabilities and limitations of machine learning models in chemistry and materials science.

benches

ChemBench

testing the chemical capabilities of llms

learn more >

MaCBench

probing vision-language models in materials science and chemistry.

learn more >

MatText

testing the limits of (l)lms in modeling materials

learn more >