© 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. Qwen3-4B
    5. 20-TRANSCODER-HP
    7. 107954
    Prev
    Next
    INDEX
    Explanations

    of

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    .anchor
    -0.26
    fig
    -0.25
    çħ®
    -0.25
    æIJĢ
    -0.25
    duk
    -0.25
     watt
    -0.25
    odon
    -0.24
    è¿ĺæĺ¯è¦ģ
    -0.24
    æĺ¯ä»¥
    -0.24
    QUE
    -0.24
    POSITIVE LOGITS
    ailles
    0.29
    pper
    0.29
    edom
    0.28
    paque
    0.27
    auty
    0.26
    verage
    0.25
    为æĤ¨
    0.24
    wers
    0.23
    ç»ĻæĤ¨
    0.23
    snake
    0.23
    Activations Density 8.401%

    No Known Activations