© 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. 55031
    Prev
    Next
    INDEX
    Explanations

    references to government accountability and public expenditure

    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
    digits
    -0.15
    oute
    -0.15
    avÄĽ
    -0.14
    Åij
    -0.14
     rád
    -0.14
    iyorum
    -0.14
    eated
    -0.14
     Evangel
    -0.14
    ãĥĸãĥª
    -0.14
     camel
    -0.14
    POSITIVE LOGITS
     ANC
    0.39
    ANC
    0.38
     South
    0.29
     Gaut
    0.28
     Johannesburg
    0.28
     Gupta
    0.28
    South
    0.27
     Esk
    0.26
     Cape
    0.25
    .za
    0.24
    Activations Density 0.039%

    No Known Activations