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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 24799
    Prev
    Next
    INDEX
    Explanations

    references to research studies and experiments involving human subjects

    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
    edor
    -0.16
     CONTRIBUTORS
    -0.14
     even
    -0.14
    esel
    -0.14
    ipherals
    -0.14
     CircularProgress
    -0.14
    à¸Ńà¸ĩà¸Ħ
    -0.14
     possibly
    -0.13
    ivor
    -0.13
     smr
    -0.13
    POSITIVE LOGITS
     varying
    0.25
     identical
    0.23
     various
    0.20
     varied
    0.20
     randomly
    0.20
     exactly
    0.19
     differing
    0.19
     either
    0.18
     then
    0.18
     each
    0.17
    Activations Density 0.064%

    No Known Activations