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

    Neuronpedia

    INDEX
    Explanations

    The neuron is looking for references to a specific location named Tah

    oai_token-act-pair · gpt-3.5-turbo

    references to "Tahrir Square" and its associated elements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    mble
    -0.74
    ISTER
    -0.71
     fixation
    -0.68
    ancial
    -0.67
    ablishment
    -0.65
     cov
    -0.65
    heid
    -0.61
    VERSION
    -0.61
     Purg
    -0.59
     quirks
    -0.59
    POSITIVE LOGITS
    oya
    1.02
    rir
    0.98
     Tah
    0.97
    anus
    0.91
    essa
    0.90
    iti
    0.89
    una
    0.88
    ania
    0.85
    ibi
    0.85
    iri
    0.85
    Act Density 0.007%

    No Known Activations