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

    Neuronpedia

    INDEX
    Explanations

    phrases related to escaping or running away from a threat

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "flee" and its variations related to escaping or leaving a situation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ammy
    -0.87
    arist
    -0.82
    essee
    -0.74
    imen
    -0.73
    ittee
    -0.68
    ocr
    -0.68
    uyomi
    -0.65
    obyl
    -0.65
    raviolet
    -0.64
    oyal
    -0.64
    POSITIVE LOGITS
     fleeing
    1.02
     flee
    0.96
     Calais
    0.77
     fled
    0.76
    ibrary
    0.68
    escape
    0.66
     abroad
    0.66
    port
    0.65
    boats
    0.64
     Torment
    0.64
    Act Density 0.011%

    No Known Activations