© 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. 28450
    Prev
    Next
    INDEX
    Explanations

    code block terminators

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

    This neuron detects the ends of code blocks (e.g. closing braces or dedent tokens signaling a function or scope closure).

    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
    komfort
    -0.95
    uksessa
    -0.92
     gewonnen
    -0.90
     verla
    -0.90
    hima
    -0.90
    我们需要
    -0.89
     Vielen
    -0.88
     их
    -0.87
     Darum
    -0.86
     verle
    -0.86
    POSITIVE LOGITS
     étudiant
    0.93
     effetto
    0.92
    equilibrium
    0.91
     empê
    0.90
    zial
    0.89
     позволя
    0.85
    Toponymie
    0.83
    źnie
    0.82
     ۱۴۰۰
    0.81
     évo
    0.81
    Activations Density 0.017%

    No Known Activations