EasyDeepLearn

Why is data deduplication critical for LLM pretraining?

medium

Answer

  • Duplicate documents cause the model to memorize verbatim rather than generalize (Carlini et al., 2020).
  • Effects: worse generalization, higher memorization risk (training data extraction attacks), inflated benchmark scores if test data leaks via duplicates, wasted compute.
  • Modern pretraining pipelines run exact-hash dedup + fuzzy dedup (MinHash / SimHash on shingles) at document and paragraph levels — often removing 20-40% of the raw corpus.
Check yourself — multiple choice
  • Dedup doesn't matter
  • Prevents memorization + benchmark leakage + wasted compute; runs exact + fuzzy (MinHash) dedup, removes 20-40% of corpus
  • Only for legal reasons
  • Only run at inference

Dedup: prevents memorization / benchmark leakage / compute waste; MinHash + exact-hash pipelines.

#pretraining#data

Practise LLMs & GenAI

214 interview questions in this topic.

Related questions