© 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
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 11990
    Prev
    Next
    INDEX
    Explanations

    code assignment

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

    This neuron activates on advertisement‐related field and parameter names (e.g. “Ad,” “Labels,” “Description,” “Path,” “Keyword,” etc.) in structured ad metadata.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_24_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     niektórych
    0.84
    𝙖
    0.81
     ಕೆಲವು
    0.80
     quaternion
    0.80
    umlahan
    0.78
     некоторых
    0.77
     dumpfile
    0.77
     někter
    0.77
    eonium
    0.75
    llabus
    0.75
    POSITIVE LOGITS
    ;
    1.26
     =
    1.16
    ;//
    0.91
    For
    0.84
     ;
    0.84
    V
    0.82
    The
    0.81
    To
    0.79
    =
    0.77
     For
    0.76
    Activations Density 0.032%

    No Known Activations