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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 31046
    Prev
    Next
    INDEX
    Explanations

    if and while loops

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

    The neuron chiefly fires on loop keywords (e.g. “for” and “while”) indicating it’s detecting loop constructs.

    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
    not
    -0.83
    arma
    -0.79
    if
    -0.79
    NOT
    -0.79
    铵
    -0.78
    ОД
    -0.77
     MER
    -0.75
     يتيمه
    -0.75
    այ
    -0.75
    eri
    -0.74
    POSITIVE LOGITS
     !(
    0.92
     (
    0.85
    ##
    0.77
     affez
    0.76
     ((
    0.75
    Apocalypse
    0.75
     preparo
    0.73
     aço
    0.70
    šem
    0.70
    Nieuwe
    0.70
    Activations Density 0.005%

    No Known Activations