© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 24242
    Prev
    Next
    INDEX
    Explanations

    the word "meant" and its various usages

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.25.hook_resid_post
    Hook Layer
    25
    Architecture
    gated
    Context Size
    1,024
    Dataset
    Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    alty
    -0.17
    sworth
    -0.14
    ansom
    -0.13
    rouch
    -0.13
     Emer
    -0.13
    .brand
    -0.13
    меÑĢикан
    -0.13
    伸
    -0.13
    ukkit
    -0.13
    elsey
    -0.13
    POSITIVE LOGITS
     to
    0.29
    ioned
    0.22
     meant
    0.20
     for
    0.20
    eken
    0.17
     intended
    0.16
    unes
    0.16
     длÑı
    0.16
     untuk
    0.16
    mods
    0.16
    Activations Density 0.022%

    No Known Activations