© 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. 22-TRANSCODER-HP
    7. 129769
    Prev
    Next
    INDEX
    Explanations

    Instructions and documentation

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    å¾Īå¤ļäºĭæĥħ
    -0.31
     interven
    -0.25
    med
    -0.25
    åĮ¯
    -0.24
    ä¸įä¸Ģå®ļ
    -0.24
     distinction
    -0.24
    ой
    -0.23
     classification
    -0.23
    $\
    -0.23
     fun
    -0.23
    POSITIVE LOGITS
    ä¸į解
    0.25
    ibbon
    0.25
    åijĬ
    0.25
    cke
    0.25
    iples
    0.25
    带æĿ¥æĽ´å¤ļ
    0.24
    ä»·æł¼ä¸Ĭ涨
    0.23
    Sibling
    0.23
    æŶéĹ´åĨħ
    0.23
    _NAMESPACE
    0.23
    Activations Density 0.002%

    No Known Activations