© 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. 40002
    Prev
    Next
    INDEX
    Explanations

    non-english characters and words

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

    The neuron flags any token that isn’t plain English text—e.g. numerals (including decimals) or non-Latin characters (Thai, Japanese, etc.).

    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
    y
    -3.27
    圌
    -3.02
    al
    -2.89
     honom
    -2.88
     моя
    -2.64
     feroit
    -2.61
    驺
    -2.61
     fantastisk
    -2.61
     stål
    -2.59
    鋱
    -2.59
    POSITIVE LOGITS
     
    3.83
     vollständige
    2.89
    но
    2.83
     frühen
    2.78
    [
    2.66
     Presently
    2.61
    1
    2.58
    ’
    2.52
    <
    2.52
    *
    2.52
    Activations Density 0.000%

    No Known Activations