© 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. 96414
    Prev
    Next
    INDEX
    Explanations

    mixed language snippets

    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
    ippy
    -0.27
    itol
    -0.27
     peak
    -0.26
    èħ»
    -0.26
     fires
    -0.26
     Fires
    -0.26
    extras
    -0.26
     Bloody
    -0.26
     rue
    -0.25
    à¹Ģย
    -0.24
    POSITIVE LOGITS
    æĬ¥
    0.28
    æĸ°çļĦ
    0.26
    ä¸įå¾Ĺ
    0.26
    oard
    0.25
    call
    0.25
    åįĹ
    0.24
    Callable
    0.24
    oser
    0.24
    SER
    0.23
    new
    0.23
    Activations Density 0.102%

    No Known Activations