RAG, Fine-Tuning & ContextNelson F. Liu et al. (Stanford, UC Berkeley, Samaya AI)Stanford UniversityUC Berkeley
Lost in the Middle published
Lost in the Middle published — a U-shaped position bias in which performance drops when the relevant document sits in the middle of the context
Why it matters
The first large-scale counterevidence to the assumption that you can simply put in a long context
Source — Primary sources first; secondary reports are labelled “Press”
Paperarxiv.orghttps://arxiv.org/abs/2307.03172Added in Week 35, 2026 · See that week