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

    Neuronpedia

    INDEX
    Explanations

    which

    np_max-act · gemini-2.0-flash

    This neuron activates on words used in explicit definitional or explanatory clauses—particularly the relative pronoun “which” and the copula “is.”

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Kok
    -0.06
     Fast
    -0.06
    <r
    -0.06
     Monument
    -0.06
     QRect
    -0.06
    Routing
    -0.06
     OSC
    -0.06
    	Title
    -0.05
    pickle
    -0.05
     metric
    -0.05
    POSITIVE LOGITS
    etes
    0.08
     فایل
    0.07
     passer
    0.07
    $")↵
    0.07
     бан
    0.07
     breathable
    0.07
    弹
    0.06
    _lua
    0.06
    Whether
    0.06
     การ
    0.06
    Act Density 0.109%

    No Known Activations