© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 104937
    Prev
    Next
    INDEX
    Explanations

    ślavic, Polish, and Hebrew prefixes

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

    The neuron fires strongly on Slavic yes–no question particles (like Czech “či” and Polish “czy”).

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
    ,"
    -2.88
    
    -2.63
    
    -2.55
     Erstellung
    -2.36
    鈊
    -2.34
     několik
    -2.34
     แต่
    -2.31
    谁
    -2.31
     ("
    -2.30
    
    -2.30
    POSITIVE LOGITS
    It
    2.59
    嫱
    2.42
    There
    2.41
    頫
    2.39
     funcion
    2.36
     beckon
    2.30
    ский
    2.28
    媜
    2.27
    ской
    2.22
    were
    2.20
    Activations Density 0.001%

    No Known Activations