© 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. 10-GEMMASCOPE-RES-131K
    7. 10648
    Prev
    Next
    INDEX
    Explanations

    percentages and cases

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

    This neuron activates on numeric tokens—counts, percentages, and other numerical data 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
     to
    -2.47
     was
    -1.95
     into
    -1.50
     from
    -1.40
     *
    -1.34
     is
    -1.30
     /*
    -1.27
     originally
    -1.23
     wasn
    -1.22
     didn
    -1.20
    POSITIVE LOGITS
     Läs
    1.44
     oreja
    1.43
     almohada
    1.42
     médicaments
    1.37
    ślę
    1.36
     zesta
    1.35
     comodidad
    1.34
     Traducción
    1.32
     muñeca
    1.30
     aidé
    1.28
    Activations Density 0.172%

    No Known Activations