Hailey Schoelkopf
  • about
  • blog (current)
  • publications
  • cv
  • Prefix Linear Attention Can Outspeed Causal Linear Attention

    Notes on Prefix Language Modeling--and a surprising observation that PrefixLM can be *faster* than Causal LM under some architectural conditions.

    12 min read   ·   August 11, 2024

    2024   ·   ML,   Architectures,   Linear-Attention,   PrefixLM

  • Linear Attention Fundamentals

    The basics of linear attention in sub-quadratic language model architectures.

    11 min read   ·   July 09, 2024

    2024   ·   ML,   Architectures,   Linear-Attention

© Copyright 2024 Hailey Schoelkopf. Powered by Jekyll with al-folio theme. Hosted by GitHub Pages. Last updated: October 30, 2024.