Model Architecture & Training MethodsPress et al.
ALiBi
AIMade by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made
ALiBi
Why it matters
Extrapolates with only a linear bias and no position embeddings; one of the two main approaches, alongside RoPE
Source — Primary sources first; secondary reports are labelled “Press”
Paperarxiv.orghttps://arxiv.org/abs/2108.12409Added in Week 35, 2026 · See that week
On the map — ⑪ Model Architecture line
- Previous stopSwitch Transformer
- This recordALiBi
- Next stopChinchilla