© 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. 10-GEMMASCOPE-RES-131K
    7. 45706
    Prev
    Next
    INDEX
    Explanations

    Russian word beginnings

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

    The neuron strongly fires on tokens written in Cyrillic (i.e. Russian words) and stays off for Latin-script tokens.

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
    1
    -3.55
    0
    -3.17
     "
    -3.03
     masterful
    -2.78
    缡
    -2.72
    izing
    -2.66
     unn
    -2.63
    ཥ
    -2.63
     unending
    -2.55
    4
    -2.55
    POSITIVE LOGITS
    獁
    3.27
     și
    3.06
    좇
    2.92
    ceda
    2.81
    摶
    2.70
     striving
    2.67
    荦
    2.66
    这是什么
    2.64
    会出现
    2.56
    逦
    2.55
    Activations Density 0.003%

    No Known Activations