icon

    Aleta Ulita

    For example, vanilla behavioral cloning on MakeWaterfall leads to an agent that strikes close to waterfalls but doesnt create waterfalls of its own, presumably because the place waterfall motion is such a tiny fraction of the actions within the demonstrations. For example, with behavioral cloning (BC), we could perform hyperparameter tuning to cut back the BC loss. Alice is successfully tuning her algorithm to the take a look at, in a means that wouldnt generalize to practical tasks, and so the 20% increase is illusory. The issue with Alice...

    Sorry, there are no results for this page.