© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 100378
    Prev
    Next
    INDEX
    Explanations

    country and region names

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

    The neuron activates on geographic place names and regional/demographic labels (e.g. states, provinces, territories, demonyms).

    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
     –
    -1.59
     the
    -1.46
     guarded
    -1.36
     "
    -1.30
     devrait
    -1.30
     casually
    -1.23
     если
    -1.16
     about
    -1.16
     c
    -1.13
    pejo
    -1.12
    POSITIVE LOGITS
     Ofer
    1.39
    mäßige
    1.38
    +"/
    1.38
    sodik
    1.33
     Testamento
    1.32
     vack
    1.32
     timmar
    1.29
    Licz
    1.29
    eloma
    1.27
    )、
    1.25
    Activations Density 0.030%

    No Known Activations