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

    Neuronpedia

    INDEX
    Explanations

    Audi

    np_max-act · gemini-2.0-flash

    This neuron detects Chinese (CJK) characters and Chinese-language tokens (non-Latin script).

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31

    garbled text

    np_max-act · gpt-5-miniTriggered by @yooniel31
    New Auto-Interp
    Negative Logits
     bahwa
    -0.07
     opravdu
    -0.07
     무엇
    -0.07
     posição
    -0.06
    ﾄ
    -0.06
    posites
    -0.06
    μβρίου
    -0.06
    čů
    -0.06
    Portály
    -0.06
     český
    -0.06
    POSITIVE LOGITS
    UNKNOWN
    0.07
     انتظ
    0.07
    CUS
    0.06
     EFF
    0.06
    cohol
    0.06
     CONSEQUENTIAL
    0.06
     Recognition
    0.06
    _DIFF
    0.06
     ":"
    0.06
    :"#
    0.06
    Act Density 19.895%

    No Known Activations