© 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. 45265
    Prev
    Next
    INDEX
    Explanations

    first appears

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

    The neuron fires strongly on boilerplate publication‐metadata phrases (e.g. “the report, which appears … online edition,” “originally appeared,” “Articles have been numbered,” etc.)

    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
     ネックレス
    -0.77
    чня
    -0.76
    ViewModels
    -0.75
    tagName
    -0.74
    بو
    -0.72
    graphics
    -0.71
     سعد
    -0.71
    glyphicon
    -0.70
    блемы
    -0.70
     Laval
    -0.69
    POSITIVE LOGITS
     appeared
    6.06
     appear
    5.69
     appearing
    5.47
     appears
    5.09
     appearance
    4.41
    appear
    4.13
    appeared
    4.06
    appearing
    4.06
    Appear
    4.00
     Appear
    3.95
    Activations Density 0.283%

    No Known Activations