© 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. 10723
    Prev
    Next
    INDEX
    Explanations

    year is, rely on, less than, place at

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

    This neuron detects tokens characteristic of generic blog‐comment spam (e.g. drug names, solicitation words, hyperlinks).

    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
     فرض
    0.94
     ام
    0.86
     Ваши
    0.80
    们的
    0.79
     أجل
    0.78
     ی
    0.77
     اغ
    0.76
    ح
    0.76
     amas
    0.74
     الكامل
    0.74
    POSITIVE LOGITS
    cing
    0.72
     cuire
    0.71
     staffed
    0.70
    ли
    0.70
     raincoat
    0.70
     cakkh
    0.68
    यदि
    0.68
    irat
    0.68
    ItemID
    0.67
    貌
    0.66
    Activations Density 0.001%

    No Known Activations