© 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. 5486
    Prev
    Next
    INDEX
    Explanations

    Italian words and phrases

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

    The neuron predominantly activates on numerical data—tokens representing numbers, percentages, measurements, or other quantitative values.

    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
     conspicuously
    0.87
    鏽
    0.84
    泄漏
    0.84
    绳
    0.82
     hurtful
    0.82
    囟
    0.82
     Беларуси
    0.81
     emphasizing
    0.80
    蛲
    0.78
     designating
    0.78
    POSITIVE LOGITS
     partecip
    1.79
     rappresent
    1.72
     grazie
    1.72
     più
    1.69
     italiani
    1.68
     indip
    1.66
     città
    1.65
     sarà
    1.63
     italian
    1.63
     abbiamo
    1.62
    Activations Density 0.346%

    No Known Activations