Dmitrii Krasheninnikov comments on Meta learning to gradient hack

[Dmitrii Krasheninnikov]

Could you please share the results in case you ended up finishing those experiments?