© 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. 18-TRANSCODER-HP
    7. 88268
    Prev
    Next
    INDEX
    Explanations

    and

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    ographed
    -0.27
     Nack
    -0.26
    aton
    -0.26
    çµIJ
    -0.26
    ç»ĵ
    -0.24
    -pic
    -0.24
    eres
    -0.23
    ivating
    -0.23
    [at
    -0.23
    erval
    -0.23
    POSITIVE LOGITS
    ä¸Ģæĸ¹éĿ¢
    0.29
    åħ¬
    0.28
    çľģå¸Ĥ
    0.27
     isot
    0.25
    ä¸Ģæĸ¹éĿ¢æĺ¯
    0.25
    :
    0.24
    è°ĥæŁ¥æĺ¾ç¤º
    0.24
    è¯Ŀ
    0.24
     intersections
    0.24
    isphere
    0.24
    Activations Density 38.513%

    No Known Activations