© 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. 24-GEMMASCOPE-2-RES-16K
    7. 11823
    Prev
    Next
    INDEX
    Explanations

    Domestic violence

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

    The neuron strongly activates on commercial service‐oriented keywords—especially titles of download‐music (“lagu”) and moving/packing services (“Movers,” “Packers,” “Removals”)—indicating it flags advertisement or directory‐style terms.

    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
    is
    0.91
    il
    0.80
    スナー
    0.79
    li
    0.78
    Royal
    0.75
    Storia
    0.73
    r
    0.73
    ir
    0.73
    é
    0.72
    ort
    0.70
    POSITIVE LOGITS
    ческую
    0.92
    ното
    0.86
    нома
    0.84
     повы
    0.82
    ческому
    0.81
     annih
    0.80
    ственной
    0.79
    穸
    0.79
     некоторых
    0.78
    йт
    0.77
    Activations Density 0.001%

    No Known Activations