{"id":"pmid:38434255","kind":"article","title":"Large language models transition from integrating across position-yoked, exponential windows to structure-yoked, power-law windows.","authors":["David Skrill","Sam Norman-Haignere"],"author_count":2,"year":2023,"published":"2023-12-01","venue":"PubMed","cited_by":2,"doi":null,"arxiv_id":null,"nct_id":null,"source_url":"https://pubmed.ncbi.nlm.nih.gov/38434255","archive_url":"https://space-kitty.com/papers/pmid:38434255","updated":null,"pmid":"38434255","pmcid":null,"openalex_id":"W4394739075","pdf_url":"https://pmc.ncbi.nlm.nih.gov/articles/PMC10907028/pdf/nihms-1969579.pdf","abstract":null,"abstract_license":null,"license":null,"is_open_access":true,"peer_reviewed":null,"categories":["Topic Modeling"],"details":null,"provenance":"openalex","attribution":"Metadata and citation counts: OpenAlex (CC0).","topics":[{"slug":"scaling-laws","name":"Scaling Laws & Compute-Optimal Training","domain_slug":"ai"}],"cited_in_dossiers":[]}