AINewsnow

They Put 7 Attention Mechanisms on a Latin Square. Then Removed Them One by One.

This story is from 2026-09-18. It is preserved in the archive; the latest stories are on the live feed.

Since GPT, nearly every Transformer repeats the same attention mechanism at every layer. Forty-eight identical blocks, differing only in learned weights. Nobody tested that. It is a convention, not a conclusion. A paper out of VIDRAFT AI Research ( arXiv:2609.20269 , CC BY 4.0) tests it, and the in…

Read the full story at DEV Community — Machine Learning ↗

Timeline · 1 report

  1. 2026-09-18 03:48 · DEV Community — Machine Learning
    They Put 7 Attention Mechanisms on a Latin Square. Then Removed Them One by One.

More stories

  1. Trump announces a new 'AI Force,' but says he will not 'stifle' AI — Business Insider AI
  2. Anthropic, OpenAI, SpaceXAI, Google sued over call to ‘pace’ AI development — Politico Technology
  3. Introducing Kimi K3 on Amazon Bedrock — AWS Machine Learning Blog
  4. Introducing Amazon SageMaker HyperPod Inference Gateway — AWS Machine Learning Blog
  5. Introducing Astra for Law — OpenAI News
  6. Microsoft exec called AI scraping the “largest theft of labor in human history” — Ars Technica AI
  7. Google Joins OpenAI, Anthropic, Meta in Disclosing AI Hacks — Bloomberg AI
  8. Sources: Anthropic considers releasing a new AI model to counter OpenAI's momentum since Astra's launch, ahead of an IPO and after Amodei's call for a slowdown (Reuters) — Techmeme

Get the daily brief of stories like this at 6:30 every morning →