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

    Neuronpedia

    INDEX
    Explanations

    Questions

    np_max-act · gemini-2.0-flash

    The neuron chiefly activates on the blank‐underscore placeholders used to denote missing words (i.e. fill‐in‐the‐blank spots) in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    inski
    -0.07
     язы
    -0.07
     Corpor
    -0.07
     comprehend
    -0.06
    kiego
    -0.06
    ніше
    -0.06
     Support
    -0.06
     Lingu
    -0.06
    获
    -0.06
     Works
    -0.06
    POSITIVE LOGITS
     democratic
    0.07
     {//
    0.06
     IBOutlet
    0.06
     [=[
    0.06
     Тем
    0.06
     encoded
    0.06
     Toastr
    0.06
    (Thread
    0.06
    (Key
    0.06
    _Number
    0.06
    Act Density 0.012%

    No Known Activations