© 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-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 10448
    Prev
    Next
    INDEX
    Explanations

    outside of acceptable content

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

    This neuron flags when an entire text segment is repeated verbatim, i.e. it detects duplicate blocks of text.

    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_31_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
    oproteins
    0.64
    黼
    0.55
     iced
    0.54
     nast
    0.54
     неба
    0.52
    prepareStatement
    0.52
     поступа
    0.52
    isecond
    0.51
    oblastic
    0.51
     callSettings
    0.51
    POSITIVE LOGITS
    Фи
    0.61
    NI
    0.60
     Strath
    0.57
     ఎమ్మె
    0.57
    эри
    0.56
     Kato
    0.55
     니
    0.55
    西北
    0.55
     Кири
    0.55
     ಎ
    0.54
    Activations Density 0.000%

    No Known Activations