For all my papers (since I am at the ANU) I publish all experimental results (of my papers accepted at top-tier venues). Note however that these ``publications'' are just snapshots of the respective code bases, scripts, and produced data (for the sake of scientific transparency). However, for all these artifacts, you should instead use the latest version of the software (or benchmarks), which should be linked both in the respective scientific paper as well as in the readme of the respective experimental repo from below.
DOIs for three of my papers/experiments have been reserved but are not yet ready: