If it was me, I would sell enough to have received a gain of 2-3x my cost basis even if the rest becomes worthless, and then just leave the rest alone unless there is a compelling reason to sell. If you aren't open to _receiving_ the truly outrageous returns that black swan-type events might cause, there is no sense in making bets for black swans. You do need to be a good sleeper to do this though.

The other criterion I'd use is that if one of my risky bets has returned enough money that I could likely retire from it (i.e. >25x annual expenses), I would sell enough to be capable of doing that. That's such a monumental quality of living change that it's not worth it to keep the same level of risk.

This. That's pretty much exactly what I did, although it was more like 800x because bitcoin. It gave me quite a bit of peace of mind to know that even if I was wrong, I was wrong in a very reasonable way and not completely out of the game.