3-subset meet-in-the-middle
A meet-in-the-middle variant that splits the key bits into three groups and works against ciphers whose key schedule merely picks out individual bits.
Classic MITM needs the cipher to fall into two halves with independent keys, as in double DES. Most modern ciphers mix the key into every round, so that rarely applies. In 2010 Andrey Bogdanov and Christian Rechberger found a way around it: partition the key into bits used only in the first half, bits used only in the second, and bits used in both. Guess the shared set, then run an ordinary MITM on the rest.
The target was KTANTAN, a lightweight RFID cipher whose key schedule does no mixing at all. The attack recovered the full key slightly faster than exhaustive search — academically marginal, but the technique became one of the building blocks of the later biclique attacks.