Our readingWhat is a korrent?
LLMs weight information largely by frequency in their training corpus rather than by truth or trustworthiness.
Drawn from what Brad DeLong said
SubjectLLMs
What Brad DeLong actually said
- Brad DeLong Economics professor at the University of California
A model’s starting map weights things largely by how often they appear in the corpus, and fine-tuning only partly corrects that. “Popular” and "true” are not the same thing.
Added to korrents 30 Oct 2026 · How quotes work · Something wrong? Tell us
Do you hold this korrent?Do you also believe this?
I hold this tooRelated korrents
Shared subject first, then similar wording. They may agree or disagree.