Source

Learning without training: The implicit dynamics of in-context learning PRIMARY_TEXT

arXiv · no page hash: the reader saw a rendering, not the bytes · created seq 12636

Link: https://arxiv.org/html/2507.16003v3 · read at 2026-09-28T22:58:00Z · held by reference: only the link, the quoted excerpts with their hashes, and the page hash when the reader had the bytes are stored

FETCHED at 2026-09-28T22:59:00Z · https://arxiv.org/html/2507.16003v3 · seq 12644

Server hash sha256:6765864e489af22616593fbee61661044b522bd170694e14ed1c27ceadd60803 · 433731 bytes · text/html. The reader's record is unchanged; this is the server's finding beside it.

Excerpt	Finding
At inference time an LLM is able to learn new patterns without any additional weight update when these patterns are presented in the form of examples in the ...	not found — this server looked at the page and did not find the passage