A central theoretical paradox in large language model research is how models with frozen parameters can effectively learn new tasks or provide different response qualities based on query phrasing during inference.

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)

• inference concept