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

    Neuronpedia

    INDEX
    Explanations

    then

    np_max-act · gemini-2.0-flash

    The neuron activates on conditional “if … then” constructs—i.e. it detects “if” statements paired with “then.”

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    贵
    -0.08
    利
    -0.07
    azard
    -0.07
    IJ
    -0.07
     mammals
    -0.06
     Labour
    -0.06
     своб
    -0.06
    �
    -0.06
    ุ์
    -0.06
    (mappedBy
    -0.06
    POSITIVE LOGITS
    Titan
    0.08
    Then
    0.07
    으면
    0.07
     Leads
    0.07
     Thom
    0.07
     sonic
    0.07
    ben
    0.07
    .then
    0.07
    čen
    0.07
     Tooth
    0.06
    Act Density 0.012%

    No Known Activations