[x] Make stimulus naming conventions consistent between model and human evaluations (Within each match scenario names to ground truth labels, @judithfan make sure these canonical names are on landing page)
[x] Create private repo for physics-benchmarking-neurips2021 and just have master branch public? @eliaszwang
physics-benchmarking-neurips2021
and just havemaster
branch public? @eliaszwang