A fundamental problem in the Data Preparation Stage of Large Language Models is quantifying how data characteristics affect model performance, including the trade-off between verbatim memorization and reasoning capabilities, the theoretical limits of synthetic data in recursive self-improvement, and the impact of data contamination on evaluation integrity.

Authors

• Person: Not available Organization: arXiv
  A Survey on the Theory and Mechanism of Large Language Models

Sources

• A Survey on the Theory and Mechanism of Large Language Models arxiv.org arXiv via serper

Referenced by nodes (1)

• Large Language Models concept