• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
  • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI
    Announcement

    Triadic linear attention proposes a three-dimensional state

    A user proposes using the outer product of three vectors to move from linear attention’s matrix state to a three-dimensional one.

    CL
    HG
    OS
    3 Sources, ,

    TLDR

    A user introduces triadic linear attention as a proposal to maintain a three-dimensional state using the outer product of three vectors. They argue that linear attention, though arguably a naive RNN, massively outperforms traditional RNNs by maintaining a matrix state.

    Combined views

    12.9K

    3 Sources, first seen 3h ago

    Combined views

    12.9K

    3 Sources, first seen 3h ago

    273 likes
    3h ago
    first seen 3h ago
    273 likes
    9 comments
    215 saves
    36 reposts
    Featured Source
    9 comments
    215 saves
    36 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    3 Sources

    @osieberlingLinear attention is arguably the most naive RNN, but still massively outperforms traditional RNNs by maintaining a matrix state. So... what if we use a (triadic) outer product of three vectors and maintain a three-dimensional state? Introducing: Triadic Linear Attention 🧵3h
    @ChengleiSiRT @osieberling: Linear attention is arguably the most naive RNN, but still massively outperforms traditional RNNs by maintaining a matrix…3h
    @HanGuo97Can we have 4D linear attention 🤔52m

    3 Sources

    @osieberlingLinear attention is arguably the most naive RNN, but still massively outperforms traditional RNNs by maintaining a matrix state. So... what if we use a (triadic) outer product of three vectors and maintain a three-dimensional state? Introducing: Triadic Linear Attention 🧵3h
    @ChengleiSiRT @osieberling: Linear attention is arguably the most naive RNN, but still massively outperforms traditional RNNs by maintaining a matrix…3h
    @HanGuo97Can we have 4D linear attention 🤔52m