© 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. 12854
    Prev
    Next
    INDEX
    Explanations

    Nepali and Malayalam scripts

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

    The neuron responds strongly to non-Latin (e.g. Indic, Burmese, Georgian, Sinhala, Thai, etc.) script characters, effectively detecting when text is written in non-Latin writing systems.

    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
     Sonos
    0.90
     súa
    0.83
     hören
    0.80
     metac
    0.79
    🏁
    0.78
    lVert
    0.78
    🍡
    0.77
     SEK
    0.77
     componente
    0.76
     ermöglicht
    0.75
    POSITIVE LOGITS
    agamanam
    0.75
    avacanam
    0.74
    akaranam
    0.68
    坳
    0.66
    த்
    0.64
     कलेक्शन
    0.64
     tatha
    0.63
    ajno
    0.63
    遑
    0.63
    峇
    0.63
    Activations Density 0.005%

    No Known Activations