© 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. 106836
    Prev
    Next
    INDEX
    Explanations

    beginning of text

    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
    æİ¥åıĹ
    -0.26
    enance
    -0.26
    æİ¥
    -0.24
    ç®±
    -0.24
    hamster
    -0.23
     convicted
    -0.23
    Vis
    -0.23
    uncture
    -0.23
    åĨĬ
    -0.23
     silent
    -0.22
    POSITIVE LOGITS
    ATING
    0.26
    èĬ±æł·
    0.25
    çİ°
    0.25
    present
    0.24
    unting
    0.24
    RAM
    0.24
    头æĿ¡
    0.24
    æµ·æ£ł
    0.23
    çīº
    0.23
    ä½Ĩçİ°åľ¨
    0.23
    Activations Density 4.814%

    No Known Activations