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

    Neuronpedia

    INDEX
    Explanations

    phrases containing the word "but" where an action or situation is obstructed or prevented

    oai_token-act-pair · gpt-3.5-turbo

    conjunctions used to introduce conflicting situations or contrasts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ocative
    -0.77
    agate
    -0.72
    conom
    -0.71
    ardon
    -0.70
    oreal
    -0.70
    onomy
    -0.68
    isable
    -0.68
    izable
    -0.68
    igmatic
    -0.67
    mortem
    -0.67
    POSITIVE LOGITS
     alas
    1.24
     luckily
    1.02
    tons
    0.94
     refused
    0.92
     fortunately
    0.92
     hesitated
    0.90
     unfortunately
    0.88
     beware
    0.86
     failed
    0.86
     stopped
    0.86
    Act Density 0.231%

    No Known Activations