© 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. 12229
    Prev
    Next
    INDEX
    Explanations

    This neuron detects terms that signal disallowed or prohibited content in guideline-style text, especially those about listing or revealing personal or identifying information.

    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
     whereupon
    0.75
     ceas
    0.73
     وق
    0.68
     परन्तु
    0.68
     postice
    0.68
     beset
    0.65
     всеми
    0.64
    opportunity
    0.64
     удовле
    0.63
     andRow
    0.63
    POSITIVE LOGITS
    比如
    0.97
     hairstyles
    0.86
     esimerkiksi
    0.86
    isetas
    0.84
    比如说
    0.83
     gimm
    0.82
     adjectives
    0.82
     hashtags
    0.80
     lyrics
    0.79
    比如說
    0.79
    Activations Density 0.017%

    No Known Activations