© 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. 25815
    Prev
    Next
    INDEX
    Explanations

    financial implications and costs associated with taxpayer expenses

    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
    aisal
    -0.15
    ymm
    -0.14
     Sleeve
    -0.14
    umps
    -0.14
    ODULE
    -0.13
    emean
    -0.13
     Tie
    -0.13
    สว
    -0.13
    coe
    -0.13
    icum
    -0.13
    POSITIVE LOGITS
     bill
    0.89
     bills
    0.69
    bill
    0.61
     Bill
    0.59
    Bill
    0.54
     Bills
    0.52
     BILL
    0.52
     tab
    0.51
    _bill
    0.50
    .bill
    0.48
    Activations Density 0.047%

    No Known Activations