© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 2710
    Prev
    Next
    INDEX
    Explanations

    preposition followed by a noun

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

    The neuron consistently lights up on numerical tokens—especially decimal numbers and percent values—i.e. it detects floating‐point or statistical numeric expressions.

    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
    marvel
    0.59
    痴
    0.55
    ได้
    0.54
    ὁ
    0.54
     карту
    0.53
    応答
    0.53
    estabelecimento
    0.53
    isValue
    0.52
    정
    0.52
    <unused13>
    0.51
    POSITIVE LOGITS
     the
    0.95
     other
    0.89
     andra
    0.73
     each
    0.71
     diğer
    0.68
     Nigeria
    0.67
     Other
    0.66
     autres
    0.65
     OTHER
    0.65
     restantes
    0.65
    Activations Density 0.268%

    No Known Activations