Skip to content
FollowtheDamnCode

How a Token Is Actually Constructed

Subword tokenization and Byte Pair Encoding — why models split words into pieces instead of reading whole words.

This lesson is part of Pro.

Free covers the C and Python courses, the first three DSA topics, the opening system design lessons and the beginner AI lessons, all with the full coach. Pro opens the rest of the library in the FollowtheDamnCode app.