© 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. 25608
    Prev
    Next
    INDEX
    Explanations

    references to partnerships and backgrounds in a financial context

    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
    FileManager
    -0.15
    gue
    -0.15
    ÐĴаж
    -0.14
    opc
    -0.14
    辺
    -0.13
    irement
    -0.13
    mentor
    -0.13
    angen
    -0.13
    .wr
    -0.12
    essler
    -0.12
    POSITIVE LOGITS
     background
    0.53
     backgrounds
    0.50
     Background
    0.42
    background
    0.41
    èĥĮæĻ¯
    0.37
    Background
    0.36
    -background
    0.35
     BACKGROUND
    0.33
     prior
    0.33
    _background
    0.32
    Activations Density 0.076%

    No Known Activations