https://www.andrew.cmu.edu/user/icaoberg/tags/nlp/ Recent content in Nlp on icaoberg Hugo en-us Sat, 08 Aug 2026 23:17:04 -0400 https://www.andrew.cmu.edu/user/icaoberg/post/2026-07-12-what-is-a-token/ Sun, 12 Jul 2026 00:00:00 +0000 https://www.andrew.cmu.edu/user/icaoberg/post/2026-07-12-what-is-a-token/ A formal definition of tokens and tokenization, and why language models process text as tokens rather than words or characters. https://www.andrew.cmu.edu/user/icaoberg/post/2026-07-10-what-is-a-rag/ Fri, 10 Jul 2026 00:00:00 +0000 https://www.andrew.cmu.edu/user/icaoberg/post/2026-07-10-what-is-a-rag/ A formal definition of Retrieval-Augmented Generation and why it is often preferable to training a language model from scratch.