© 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. 100168
    Prev
    Next
    INDEX
    Explanations

    graphics elements

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

    This neuron fires on tokens naming structural elements or types (e.g. SVG tags like “g” or “path,” class/constructor names like “MovieClip” or “Group,” and similar element/type identifiers).

    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
    in
    -1.48
     الثالث
    -1.41
     otros
    -1.39
     Gespräche
    -1.38
    也在
    -1.38
     last
    -1.30
    owników
    -1.29
    #
    -1.28
     third
    -1.28
    ''
    -1.27
    POSITIVE LOGITS
    繍
    1.56
    ？！
    1.48
    ják
    1.39
    fungsi
    1.36
    詭
    1.33
    合作伙伴
    1.30
    Batis
    1.29
    ープン
    1.27
     coste
    1.26
    biblioteca
    1.26
    Activations Density 0.005%

    No Known Activations