© 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. 107395
    Prev
    Next
    INDEX
    Explanations

    Microsoft

    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
    estroy
    -0.28
    çī¹åĪ«å£°æĺİ
    -0.26
    ahead
    -0.26
    .extension
    -0.25
    abouts
    -0.25
    peria
    -0.25
     beforehand
    -0.24
    thing
    -0.24
    ä¸Ĭåİ»
    -0.23
    #----------------------------------------------------------------
    -0.23
    POSITIVE LOGITS
    s
    0.32
     rhythm
    0.27
    sı
    0.26
     syll
    0.25
    зван
    0.25
    é¦ĸéĥ½
    0.25
    ertz
    0.25
    sheets
    0.24
    à¸IJาà¸Ļ
    0.24
    ullen
    0.23
    Activations Density 11.907%

    No Known Activations