© 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. 31-GEMMASCOPE-2-RES-16K
    7. 1791
    Prev
    Next
    INDEX
    Explanations

    widespread and waking

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

    This neuron detects tokens that begin with a capital “W” followed by lowercase letters (e.g. proper names or brands starting with “W”).

    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_31_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
    ocent
    0.59
    ventura
    0.57
     질
    0.55
     Bhav
    0.54
     OSError
    0.54
     Etappe
    0.53
     Festa
    0.53
    ovic
    0.53
     Inoltre
    0.53
     Deze
    0.52
    POSITIVE LOGITS
    idespread
    1.18
     W
    0.97
    illiams
    0.95
    ounded
    0.87
    imbledon
    0.86
    W
    0.86
    avelength
    0.84
    w
    0.84
    WI
    0.83
    elfare
    0.82
    Activations Density 0.422%

    No Known Activations