© 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. 30-TRANSCODER-HP
    7. 96567
    Prev
    Next
    INDEX
    Explanations

    bang

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åĬ³
    -0.33
    .dds
    -0.29
    åĭŀ
    -0.28
    åĬ³åĬ¨
    -0.28
    PORT
    -0.27
    .Parcel
    -0.27
    fühl
    -0.27
     GLenum
    -0.26
    sns
    -0.26
    ald
    -0.26
    POSITIVE LOGITS
    .getLine
    0.28
     pan
    0.27
     line
    0.25
    æĺµ
    0.24
    å¾ĢåIJİ
    0.24
    ÑĤиÑĢ
    0.24
     orderly
    0.23
    梨
    0.23
    mie
    0.23
    tring
    0.23
    Activations Density 0.038%

    No Known Activations