© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 90463
    Prev
    Next
    INDEX
    Explanations

    mathematical and physics equations

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

    The neuron detects LaTeX‐style mathematical notation (e.g., backslashes, “frac,” “left,” “right,” parentheses, and other math formatting tokens).

    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
    诧异
    -1.76
     to
    -1.71
    ︲
    -1.68
     schöne
    -1.67
     другие
    -1.64
     Pedido
    -1.64
    /
    -1.59
     andere
    -1.58
     
    -1.58
    轳
    -1.54
    POSITIVE LOGITS
    ׇ
    1.98
     pertandingan
    1.91
     tertinggi
    1.87
     چربی
    1.83
     lapisan
    1.81
     our
    1.79
    Despite
    1.77
     gelar
    1.76
     lakukan
    1.73
    ࣭
    1.72
    Activations Density 0.020%

    No Known Activations