© 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. 10-GEMMASCOPE-RES-131K
    7. 11674
    Prev
    Next
    INDEX
    Explanations

    quotation marks and programming delimiters

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

    opening quotation marks, especially at the start of quoted strings or code snippets.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron is especially responsive to quote‐mark characters (single or double quotation marks) in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    听到了
    -1.65
     was
    -1.60
     -
    -1.59
    Another
    -1.56
     motivadoras
    -1.55
    があります
    -1.54
     gainera
    -1.54
    !
    -1.51
    That
    -1.51
     invented
    -1.46
    POSITIVE LOGITS
     äußer
    1.74
    竽
    1.59
     were
    1.54
     соответствии
    1.48
    issent
    1.45
    they
    1.41
    enaar
    1.40
     consectetur
    1.39
    GillaGilla
    1.38
     Tapi
    1.38
    Activations Density 0.007%

    No Known Activations