© 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. 34-GEMMASCOPE-RES-131K
    7. 33496
    Prev
    Next
    INDEX
    Explanations

    C++ preprocessor directives

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

    The neuron activates on document‐level formatting delimiters—e.g. horizontal rules (lines of “---”), backslash markers, code‐block fences and other section‐separator syntax.

    oai_token-act-pair · o4-miniTriggered 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
    vfill
    -0.97
    typeout
    -0.93
    begingroup
    -0.88
    clearpage
    -0.88
    liced
    -0.87
    textnormal
    -0.83
    黑白
    -0.82
    smallskip
    -0.82
    uitable
    -0.80
    沌
    -0.79
    POSITIVE LOGITS
    begin
    1.13
     paragraph
    0.94
    label
    0.91
    !\
    0.89
     paragraphs
    0.85
     label
    0.83
    rillo
    0.83
    paragraph
    0.82
     \\
    0.81
     refurbished
    0.79
    Activations Density 0.001%

    No Known Activations