© 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. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 47405
    Prev
    Next
    INDEX
    Explanations

    words starting with ze

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron consistently activates on the character sequence “Ze” at the start of words (e.g. Zeeman, Zelig, Zeigler, Zearalenone).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    อยู่
    -0.79
     deste
    -0.74
    const
    -0.70
    اق
    -0.69
     recovers
    -0.69
     Rost
    -0.68
     увидеть
    -0.67
     droppings
    -0.67
     unfortunately
    -0.66
    acon
    -0.66
    POSITIVE LOGITS
    Ze
    0.93
     Ze
    0.90
    zé
    0.82
    َي
    0.81
     ze
    0.81
     Зе
    0.79
    řejmě
    0.78
    XY
    0.75
    𖣔
    0.74
     belki
    0.73
    Activations Density 0.019%

    No Known Activations