Address clustering heuristic in BlockSci

[Natkamon]

I would like to confirm with the BlockSci paper about what the heuristics do BlockSci use for address clustering?

Does the address clustering also detect mixing service to reduce the false clusters?