© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 6115
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on non-English subword fragments—particularly pieces of words from foreign languages (e.g. Gaelic or other accented/compound tokens).

    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_24_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
    是一位
    0.66
    是最
    0.59
     является
    0.58
    भाजपा
    0.57
     rö
    0.53
    是一名
    0.52
     П
    0.50
    是一個
    0.50
     dobb
    0.50
     irrevoc
    0.50
    POSITIVE LOGITS
    odenum
    0.71
     Mereka
    0.64
    ippings
    0.64
     Nurses
    0.64
     dimana
    0.63
     Manajemen
    0.63
    ر
    0.63
     theſe
    0.62
     که
    0.62
    ressing
    0.62
    Activations Density 0.005%

    No Known Activations