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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating a positive evaluation or outcome

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating potential or expectations of outcomes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    soever
    -0.79
    phia
    -0.69
     Ples
    -0.64
     quar
    -0.62
     moreover
    -0.61
     meanwhile
    -0.60
    started
    -0.59
     Steps
    -0.58
     secondly
    -0.58
     ceased
    -0.57
    POSITIVE LOGITS
     compelling
    0.78
    risome
    0.77
    bidden
    0.75
    cellent
    0.72
    uci
    0.72
     interesting
    0.71
    asty
    0.70
     awkward
    0.69
     tidy
    0.69
    avorable
    0.69
    Act Density 0.065%

    No Known Activations