Phil 2.9.21

Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention

  • Transformers have emerged as a powerful tool for a broad range of natural language processing tasks. A key component that drives the impressive performance of Transformers is the self-attention mechanism that encodes the influence or dependence of other tokens on each specific token. While beneficial, the quadratic complexity of self-attention on the input sequence length has limited its application to longer sequences — a topic being actively studied in the community. To address this limitation, we propose Nyströmformer — a model that exhibits favorable scalability as a function of sequence length. Our idea is based on adapting the Nyström method to approximate standard self-attention with O(n) complexity. The scalability of Nyströmformer enables application to longer sequences with thousands of tokens. We perform evaluations on multiple downstream tasks on the GLUE benchmark and IMDB reviews with standard sequence length, and find that our Nyströmformer performs comparably, or in a few cases, even slightly better, than standard Transformer. Our code is at this https URL.

Book – Working on snippets

GPT Agents

  • 10:00 Meeting with Jay Alammar – that went *really* well!
  • 3:00 Meeting that I’m going to be a bit late for


  • 11:00 Meeting with Vadim
  • Nothing obvious about the roll problems, but that shouldn’t stop the work on getting the pitch maneuver running and generating data
    • Add rwheel efficiencies to the script in some kind of loop
    • Adjust AngleController to use rwheel efficiency from the ddict
  • Got my intersection code running well
  • 10:00 Meeting for first period report. Hopefully we’ll be able to get a meeting with the TPOC next week to see what he wants?