© 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. 31-TRANSCODER-HP
    7. 17949
    Prev
    Next
    INDEX
    Explanations

    center

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    gu
    -0.26
    hiba
    -0.25
    åİī
    -0.25
    ç¦Ħ
    -0.24
    wal
    -0.24
    oven
    -0.24
    onal
    -0.24
    ollection
    -0.24
    kah
    -0.24
    kad
    -0.23
    POSITIVE LOGITS
     cent
    0.77
    ä¸Ńå¿ĥ
    0.76
     center
    0.76
    cent
    0.72
     centers
    0.71
     CENT
    0.69
     ä¸Ń
    0.68
    center
    0.68
    -cent
    0.68
     central
    0.68
    Activations Density 35.775%

    No Known Activations