© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 66122
    Prev
    Next
    INDEX
    Explanations

    specific headings or questions

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

    The neuron detects numeric literals (sequences of digits) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    segments of source code—especially WeChat Mini Program (WXML/WXSS) markup and related JavaScript/Node.js module code.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    whitespace and indentation in code, particularly at the beginning of lines in structured markup and programming languages.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    ادر
    -0.80
     العلاقات
    -0.75
     pourriez
    -0.75
     Wesen
    -0.72
     \%}$
    -0.71
     relaciones
    -0.71
     به
    -0.70
    ̚
    -0.69
    口的
    -0.68
    ):
    -0.67
    POSITIVE LOGITS
    wx
    1.50
     wx
    1.32
    getApp
    1.06
     getApp
    1.05
    console
    1.00
    var
    0.99
     app
    0.92
     App
    0.90
     shale
    0.88
     that
    0.86
    Activations Density 0.005%

    No Known Activations