© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 1644
    Prev
    Next
    INDEX
    Explanations

    U.S. state of

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

    The neuron activates on the common administrative‐location phrases in encyclopedia entries—i.e. tokens like “is a [type] in the… of the…,” including division labels (town, district, county, state) and hierarchy words (“in,” “of,” “the”).

    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
    lloworld
    -1.00
     MNRAS
    -0.92
     this
    -0.91
     órdenes
    -0.90
     muualla
    -0.88
    ченной
    -0.88
    olese
    -0.88
    entibus
    -0.87
     varied
    -0.86
    際には
    -0.86
    POSITIVE LOGITS
     in
    1.25
     российского
    0.97
     it
    0.97
    gued
    0.94
     нужен
    0.91
     located
    0.90
     типов
    0.90
     of
    0.90
     входит
    0.90
     pertenece
    0.85
    Activations Density 0.196%

    No Known Activations