© 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. 84207
    Prev
    Next
    INDEX
    Explanations

    quotation followed by speech verbs

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

    The neuron flags the tokens that mark or introduce direct speech—i.e. the beginnings of quoted dialogue and its speaker attribution.

    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
     Allí
    -1.06
    Incluye
    -0.99
     seguintes
    -0.97
    ấc
    -0.94
    drücke
    -0.93
    layah
    -0.92
     绘
    -0.91
    週末
    -0.91
     $$$
    -0.91
    뷔
    -0.91
    POSITIVE LOGITS
    流石
    0.97
    nement
    0.89
     Чи
    0.88
    İM
    0.88
     whilst
    0.85
     dichiarato
    0.85
     stated
    0.83
     Ö
    0.83
     priori
    0.82
    件事情
    0.82
    Activations Density 0.013%

    No Known Activations