Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    ocean basin

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

    the neuron detects tokens belonging to non-Latin / foreign-language text (e.g., Cyrillic or other non-English script segments).

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31
    New Auto-Interp
    Negative Logits
     ases
    0.35
     degrade
    0.33
     airbags
    0.31
     reduzir
    0.31
     grinder
    0.31
     magnets
    0.30
     atualizar
    0.30
     market
    0.29
    डेट
    0.29
    ικο
    0.29
    POSITIVE LOGITS
     Furthermore
    0.29
     Moreover
    0.29
     Scienze
    0.28
    ↵
    0.28
    Finally
    0.27
     Finally
    0.27
     Paragraph
    0.26
    សូម
    0.26
     Finalmente
    0.26
     Merriam
    0.26
    Act Density 0.346%

    No Known Activations