© 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. 129467
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on title‐like lines or headings—sequences of capitalized words and enclosing punctuation (e.g. parentheses) that mark titles or section headings.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    translations and foreign text

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    ;:
    -1.00
     aye
    -0.88
    gewiesen
    -0.88
     arriv
    -0.86
    .
    -0.84
    "}>
    -0.83
     spek
    -0.79
    estias
    -0.78
    uertos
    -0.77
     paloma
    -0.76
    POSITIVE LOGITS
     šķ
    0.97
    مقدمه
    0.94
     translation
    0.93
    鷺
    0.90
     translated
    0.89
    Translation
    0.87
     english
    0.86
    印度
    0.85
    🎽
    0.84
    leste
    0.83
    Activations Density 0.060%

    No Known Activations