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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 65902
    Prev
    Next
    INDEX
    Explanations

    quotation marks and structure delimiters

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

    The neuron is detecting long runs of leading whitespace (indentation) at the start of lines.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     mikina
    -1.98
    Jeśli
    -1.89
    Zubereitung
    -1.81
    ☫
    -1.80
     Drž
    -1.80
     if
    -1.80
     biela
    -1.80
     fasse
    -1.79
     対策
    -1.78
    moiselle
    -1.77
    POSITIVE LOGITS
     “
    2.73
     ‘
    2.05
    _
    1.95
    ’
    1.88
     ____________
    1.77
    '
    1.73
    >
    1.73
    Y
    1.71
    k
    1.66
    {
    1.64
    Activations Density 0.032%

    No Known Activations