Thread

1. bilater · 1789842804 · focus · HN ↗

   I wonder if there is a principle here we can apply to neural nets like dividing up the layers for different tasks (no not MoE but something more fundamental).

   1. AWTom · 1789845424 · focus · HN ↗

      There is some research on layer duplication on existing pre-trained models. This paper also references layer skipping: <a href="https:&#x2F;&#x2F;aclanthology.org&#x2F;2025.findings-emnlp.967.pdf" rel="nofollow">https:&#x2F;&#x2F;aclanthology.org&#x2F;2025.findings-emnlp.967.pdf

      1. bilater · 1789929324 · focus · HN ↗

         nice

Unofficial Hacker News client; not affiliated with Y Combinator.