© 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. 117456
    Prev
    Next
    INDEX
    Explanations

    code definitions and initializations

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

    The neuron strongly activates on C/C++ style data‐declaration tokens (e.g. “const uint8_t identifier[…]”) and thus is detecting code declarations.

    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
     pů
    -0.77
    ੈ
    -0.73
    pita
    -0.72
     sembra
    -0.71
    tnt
    -0.71
     unterschiedlich
    -0.70
     лучше
    -0.70
     августа
    -0.69
    hWnd
    -0.69
     szá
    -0.68
    POSITIVE LOGITS
     table
    1.05
     табли
    1.04
    ={
    1.00
     seznam
    0.95
    table
    0.91
     también
    0.91
     also
    0.90
    mtable
    0.88
    setDescription
    0.88
    ardı
    0.88
    Activations Density 0.040%

    No Known Activations