© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 20-TRANSCODER-HP
    7. 142761
    Prev
    Next
    INDEX
    Explanations

    Punctuation and "and"

    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
    Negative Logits
    oce
    -0.26
    dling
    -0.26
    峪
    -0.25
    unread
    -0.25
    ợi
    -0.24
    pare
    -0.24
     culo
    -0.24
    IÅŀ
    -0.23
     zwiÄħzane
    -0.23
    THEN
    -0.23
    POSITIVE LOGITS
    æľīçļĦ
    0.28
     STILL
    0.28
    æ±Ģ
    0.27
     Sik
    0.26
    è¿ĺä¸į
    0.25
    åĵĪåĵĪ
    0.25
     generally
    0.24
    ç«Ł
    0.24
    ccion
    0.24
    åľ»
    0.24
    Activations Density 10.654%

    No Known Activations