© 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. 12588
    Prev
    Next
    INDEX
    Explanations

    jobClonegreedyownership

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

    This neuron fires strongly at tokens that begin new document or advertisement‐style sections (i.e. the first word or URL fragment after a blank line), effectively marking section or topic boundaries.

    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
    เรา
    0.86
     attiv
    0.78
    mesi
    0.77
    चलिए
    0.77
     adulti
    0.75
    ’-
    0.73
    คุณ
    0.72
     suffit
    0.71
     raccol
    0.71
    }}}$
    0.70
    POSITIVE LOGITS
     сча
    0.81
    еты
    0.80
     diferentes
    0.80
     должности
    0.79
     неболь
    0.77
     перевоз
    0.77
     hemodynamic
    0.75
     desempen
    0.73
     пациентов
    0.73
     decarbon
    0.73
    Activations Density 0.001%

    No Known Activations