Mixed training creates more shared gradient-influenced parameters between fact-storing and fact-answering than two-stage training, and these shared parameters are what enable generalization to unseen questions.
Physics of language models: Part 3.2, knowledge manipula- tion
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Understanding Fact Recall in Language Models: Why Two-Stage Training Encourages Memorization but Mixed Training Teaches Knowledge
Mixed training creates more shared gradient-influenced parameters between fact-storing and fact-answering than two-stage training, and these shared parameters are what enable generalization to unseen questions.