The 4-Look Last Layer involves completing the last layer in four steps or looks. It is an intermediate step between the beginner LBL method and learning full Fridrich. It is generally broken down into 2-look OLL (orientation of edges, followed by orientation of corners) and 2-look PLL (permutation of corners, followed by permutation of edges). Achieving 4LLL requires knowledge of 16 algorithms or 14 excluding mirrors. While progressing from 4LLL to 2LLL a speedcuber will typically learn all 21 PLLs first, to achieve a 3-look last layer. Eventually, learning the final 57 OLL cases will achieve the 2-look last layer used in full Fridrich.
Another possible breakdown is to use 2-look CLL (CO-CP) and 2-look ELL (EO-EP). This can later become three different 3-looks, full CLL and 2-look ELL, 2-look CLL and full ELL but also CO, CPEOLL and EPLL. This suggestion only uses 26 algorithms in total, the drawback is that the mid step is not of much use when expanding to 2-look.
Both the 2-Look CLL and ELLs can also be done with permutations before orientations. This takes learning 14 algorithms (or 16 if including mirrored algorithms). Splitting CLL into 4-look reduces the algorithms needed for permutation from 2 to 1, and the number needed for corner orientation to 1 (or 2 including its mirror). Splitting ELL into 4-look reduces the number of required algorithms for permutation from 4 to 1 (or 2 including its mirror) and reduces the permutation to 1 algorithm.
Another breakdown is BLL, which uses a 2-Look LLEF and 2-Look CLL (OCLL-EPP - CPLL) that respects the edges. This can develop into a 3LLL by perform the LLEF in one step. This 3LLL method uses only 24 algorithms (26 including reuses), and can be later expanded to a 2LLL that uses 98 algorithms (exactly 100 including reuses).