© 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. 108657
    Prev
    Next
    INDEX
    Explanations

    Filling/levels

    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
    ç±»åŀĭ
    -0.26
     intellect
    -0.26
    ija
    -0.25
     STILL
    -0.24
     ep
    -0.24
     testim
    -0.24
    许åı¯
    -0.24
    许åı¯è¯ģ
    -0.23
    ç¬ı
    -0.23
     gå
    -0.23
    POSITIVE LOGITS
     mainland
    0.29
     conj
    0.28
    unan
    0.26
    [..
    0.25
     Conj
    0.25
    +"'
    0.25
    conj
    0.24
     fault
    0.23
    odox
    0.23
     Impact
    0.23
    Activations Density 2.203%

    No Known Activations