© 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. 5294
    Prev
    Next
    INDEX
    Explanations

    كتير حلو

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

    This neuron activates on isolated non-English or non-Latin tokens—especially numerals and foreign-script (e.g. Arabic/Persian) fragments—marking parts of the text that fall outside basic English words.

    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
    R
    0.68
    r
    0.67
    ;
    0.67
    and
    0.65
    S
    0.64
    IT
    0.61
    H
    0.61
    Ọ
    0.61
    Pokémon
    0.61
    J
    0.61
    POSITIVE LOGITS
    ってる
    1.09
     اك
    0.97
    ينه
    0.93
     مثل
    0.91
    يك
    0.91
    してる
    0.91
    ونا
    0.88
     اذا
    0.88
     اون
    0.88
     اللي
    0.88
    Activations Density 0.009%

    No Known Activations