© 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. 34-GEMMASCOPE-RES-131K
    7. 102525
    Prev
    Next
    INDEX
    Explanations

    Świątniki, Buddhist śīla, code comments, This project

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

    The neuron is highly sensitive to Polish diacritic characters (e.g. Ś, ą, ę) in tokens, especially those forming place names like “Świątniki.”

    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
    yla
    -0.80
     LCCN
    -0.78
     எ
    -0.73
    tapes
    -0.70
     Gabriel
    -0.69
    build
    -0.69
    arming
    -0.69
    обе
    -0.68
     Quy
    -0.68
     RM
    -0.68
    POSITIVE LOGITS
     inteligen
    0.87
    }^{-\
    0.81
    variables
    0.81
    ėje
    0.79
    ~!
    0.78
     punishable
    0.77
     lidé
    0.76
     چیز
    0.74
     composto
    0.74
     yours
    0.73
    Activations Density 0.014%

    No Known Activations