© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 3531
    Prev
    Next
    INDEX
    Explanations

    foreign language initiators

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

    The neuron fires on words that directly address the reader or set up instructions—chiefly the second‐person “you/you’ll” forms and related conditional or connective cues (if, when, once, should).

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
    0.81
     ubiquitous
    0.80
    s
    0.77
    /
    0.73
     -
    0.70
     support
    0.70
     campaign
    0.69
     ardından
    0.68
     simulations
    0.67
     unofficial
    0.67
    POSITIVE LOGITS
     Tetapi
    0.83
     সমস্ত
    0.74
    ientras
    0.73
     Entonces
    0.71
    ⅽ
    0.70
    newNode
    0.70
    沢山
    0.70
    quanto
    0.69
    પણ
    0.69
    ecause
    0.68
    Activations Density 0.664%

    No Known Activations