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

    Neuronpedia

    INDEX
    Explanations

    is

    np_max-act · gemini-2.0-flash

    The neuron appears never to fire (all activation scores are zero), i.e. it’s effectively a “dead” neuron with no identifiable pattern it responds to.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     floral
    -0.06
     Тут
    -0.06
    booking
    -0.06
    _successful
    -0.06
    .attack
    -0.06
    anooga
    -0.06
    [Unit
    -0.06
     Teams
    -0.06
     Gives
    -0.06
     lent
    -0.06
    POSITIVE LOGITS
     tener
    0.06
    どこ
    0.06
     reinforces
    0.06
     buffers
    0.06
     os
    0.06
     PendingIntent
    0.06
     winners
    0.06
    assertEquals
    0.06
    ,ev
    0.06
     mattered
    0.05
    Act Density 0.008%

    No Known Activations