© 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. 130069
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    ç»ıåħ¸
    -0.28
     assert
    -0.25
    çĽijäºĭä¼ļ
    -0.24
    à¸Ńม
    -0.24
    .attrs
    -0.24
     slee
    -0.24
     appointment
    -0.23
     pot
    -0.23
     warranty
    -0.23
    ework
    -0.23
    POSITIVE LOGITS
     Vent
    0.31
    zan
    0.28
    yan
    0.27
    æ°§æ°Ķ
    0.25
    çIJ¢ç£¨
    0.25
    æ®ĭéħ·
    0.25
    è¿İæİ¥
    0.25
    inja
    0.24
    _arguments
    0.24
    å£ķ
    0.24
    Activations Density 1.116%

    No Known Activations

    This feature has no known activations.