© 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. 14850
    Prev
    Next
    INDEX
    Explanations

    Mio, Oia, Erik, asio, Mondial

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

    This neuron detects isolated non-Latin characters (e.g. single Greek or Japanese letters) 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
    不会
    -2.64
    <bos>
    -2.52
    不能
    -2.42
    前は
    -2.41
     δεν
    -2.31
     we
    -2.20
    u
    -2.17
     But
    -2.14
    quelize
    -2.14
    後は
    -2.13
    POSITIVE LOGITS
    辚
    2.50
    \
    2.44
    tedì
    2.39
    راین
    2.31
     جانب
    2.30
    霎
    2.28
    ⦾
    2.28
    翃
    2.28
     consequence
    2.25
    翆
    2.25
    Activations Density 0.000%

    No Known Activations