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

    Neuronpedia

    INDEX
    Explanations

    phrases conveying a condition followed by an alternative outcome or scenario

    oai_token-act-pair · gpt-3.5-turbo

    phrases emphasizing conditional statements and the idea of believing or accepting different viewpoints despite contradictions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    onement
    -0.69
    NetMessage
    -0.68
    udding
    -0.63
    .............
    -0.61
     Eg
    -0.60
     âĶľâĶĢâĶĢ
    -0.60
    resa
    -0.60
     Reply
    -0.58
    âĸij
    -0.58
    ONES
    -0.57
    POSITIVE LOGITS
     technically
    1.07
     admittedly
    0.83
     ostensibly
    0.78
     physically
    0.77
     otherwise
    0.75
     trivial
    0.72
     outward
    0.71
     concede
    0.69
     occasional
    0.69
     admit
    0.69
    Act Density 0.307%

    No Known Activations