© 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-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 5130
    Prev
    Next
    INDEX
    Explanations

    actions that lead to

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

    This neuron fires on mentions of “Daten” or “Informationen” (especially in the context of personal data and data‐protection language).

    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
    时间和
    0.60
    妸
    0.59
    끅
    0.58
     करेगी
    0.54
    мые
    0.53
    荏
    0.52
    雱
    0.52
    땐
    0.52
    需要的
    0.50
    招标文件
    0.50
    POSITIVE LOGITS
     (
    0.64
     sowie
    0.63
     (‘
    0.61
     以及
    0.58
     which
    0.58
     ("
    0.57
    以及
    0.57
    ։
    0.56
     (**
    0.55
     (+
    0.54
    Activations Density 0.002%

    No Known Activations