© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 7008
    Prev
    Next
    INDEX
    Explanations

    Su- prefix fighter jets, suggestive

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

    The neuron consistently activates on tokens beginning with “Su” (e.g. Su, Sucre, Sundar, Susie).

    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
    ⃘
    0.57
     samos
    0.55
     segi
    0.55
     sawing
    0.54
    
    0.54
    acht
    0.53
    ecção
    0.53
    ☄
    0.52
     sns
    0.51
    bles
    0.50
    POSITIVE LOGITS
    Su
    1.06
     Su
    0.97
     Су
    0.93
     Sup
    0.90
    Sup
    0.89
    苏
    0.88
    Су
    0.86
     Sud
    0.84
     су
    0.81
     Sul
    0.79
    Activations Density 0.401%

    No Known Activations