© 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. 22-GEMMASCOPE-RES-131K
    7. 36713
    Prev
    Next
    INDEX
    Explanations

    closing code blocks

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

    The neuron fires on code tokens that look like executable statements—especially function or method calls (including the dot, call name, parentheses, semicolon) and similar imperative keywords.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     if
    -0.97
     гемо
    -0.95
     prêtres
    -0.87
     depends
    -0.82
     алгорит
    -0.81
     компре
    -0.77
     пато
    -0.76
     Bowls
    -0.75
     jeśli
    -0.74
    ましょう
    -0.73
    POSITIVE LOGITS
    OPE
    0.93
     RB
    0.91
     Japanese
    0.90
    𝒲
    0.89
    EREF
    0.85
     работы
    0.85
     Fab
    0.84
    椒
    0.84
     Associated
    0.83
     Romeo
    0.83
    Activations Density 0.001%

    No Known Activations