© 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. 6685
    Prev
    Next
    INDEX
    Explanations

    lossy, j, charge

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

    The neuron is highly sensitive to numeric tokens—years, page numbers, measurements, and other digit sequences.

    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
    ності
    0.69
    ueling
    0.68
     Immunity
    0.66
    های
    0.65
    iers
    0.64
    ifikasi
    0.63
    '
    0.61
    ної
    0.61
     হোসেন
    0.61
    फा
    0.60
    POSITIVE LOGITS
     konto
    0.82
    ᅬ
    0.82
    AMINATION
    0.77
    ExprNode
    0.75
    ნდა
    0.75
    Pg
    0.75
     små
    0.75
     ktoś
    0.74
     contemplates
    0.74
     frecuentes
    0.73
    Activations Density 0.000%

    No Known Activations