"Faustino da Silva et al.": https://ieeexplore.ieee.org/document/9370322
Might, e.g., be interesting for evaluating solvers or techniques to gain precision on a large scale.
I came across it because of an ASE paper by CEA List where they used it to learn loop unrolling heuristics (https://cea.hal.science/cea-05394085/file/2025347559.pdf)