© 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. 25123
    Prev
    Next
    INDEX
    Explanations

    references to user experience and interaction with products or services

    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
    only
    -0.21
     only
    -0.20
     also
    -0.16
     ONLY
    -0.16
     entire
    -0.16
     not
    -0.15
    -only
    -0.15
     still
    -0.15
    lut
    -0.15
     another
    -0.14
    POSITIVE LOGITS
     chose
    0.23
     choose
    0.20
     ultimately
    0.20
     chosen
    0.19
     most
    0.18
     chooses
    0.18
     MOST
    0.17
    choose
    0.17
    Choose
    0.17
    currently
    0.17
    Activations Density 0.132%

    No Known Activations