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

    Neuronpedia

    INDEX
    Explanations

    specific clauses starting with "be it"

    oai_token-act-pair · gpt-3.5-turbo

    phrasing that emphasizes identity or existence

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ogun
    -0.70
     plates
    -0.67
    ETA
    -0.66
     tape
    -0.63
    inav
    -0.62
    Ring
    -0.62
    redit
    -0.61
    uda
    -0.60
    uracy
    -0.60
    Uk
    -0.60
    POSITIVE LOGITS
    nia
    0.75
    adin
    0.63
     construed
    0.63
     srf
    0.62
    lehem
    0.61
     ç¥ŀ
    0.59
     remotely
    0.58
     prec
    0.57
     Incarn
    0.56
    hemat
    0.56
    Act Density 0.050%

    No Known Activations