© 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. 22-GEMMASCOPE-RES-131K
    7. 75078
    Prev
    Next
    INDEX
    Explanations

    code separators

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

    The neuron fires on lines made up of long runs of punctuation characters (e.g. repeated slashes or asterisks) used as decorative comment separators.

    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
    鰱
    -0.97
     コンパクト
    -0.91
     būt
    -0.90
    ظه
    -0.90
    ᨆ
    -0.89
    发布时间
    -0.88
    ピール
    -0.88
    ashing
    -0.88
     Pada
    -0.87
    俾
    -0.87
    POSITIVE LOGITS
     المتح
    1.05
     miscellaneous
    0.98
    接下来
    0.97
     порядка
    0.94
    !
    0.90
     аллер
    0.90
    forName
    0.89
    ⏟
    0.89
    てもら
    0.88
    很是
    0.88
    Activations Density 0.024%

    No Known Activations