© 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. 73864
    Prev
    Next
    INDEX
    Explanations

    English and foreign words

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

    This neuron selectively responds to structural metadata and label tokens—e.g. tag names, field names, dates, units, and other label-like words in XML/PLIST headers.

    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
    yaratan
    -0.86
    чна
    -0.73
     LYON
    -0.71
     Lyon
    -0.70
     reject
    -0.69
    シル
    -0.69
    흉
    -0.69
    buzo
    -0.69
    查询
    -0.68
     replication
    -0.68
    POSITIVE LOGITS
     accel
    0.93
     anonym
    0.79
    replaceAll
    0.77
     Asser
    0.77
     acciones
    0.76
    mentos
    0.76
     pantallas
    0.75
    安く
    0.74
    PROBLE
    0.74
     автомо
    0.74
    Activations Density 0.006%

    No Known Activations