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

    Neuronpedia

    INDEX
    Explanations

    conditions for life

    np_max-act · gemini-2.0-flash

    This neuron detects ordinary text tokens in the assistant’s generated answer passages (i.e. non‐header, non‐control words in the assistant’s prose).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    буд
    -0.07
     下
    -0.07
    Interop
    -0.07
     预
    -0.07
    IST
    -0.06
    下
    -0.06
    _app
    -0.06
    Color
    -0.06
    壁
    -0.06
    sector
    -0.06
    POSITIVE LOGITS
    (simp
    0.07
     몰
    0.07
    ighted
    0.07
    rix
    0.06
    virt
    0.06
    ترنت
    0.06
     Slack
    0.06
     Bahrain
    0.06
    jes
    0.06
    렉
    0.06
    Act Density 0.013%

    No Known Activations