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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating coercion or forceful actions

    oai_token-act-pair · gpt-3.5-turbo

    instances of being compelled or forced to take specific actions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    laughter
    -0.72
     Compass
    -0.66
     Blueprint
    -0.66
    icious
    -0.65
     Prediction
    -0.65
     Garland
    -0.64
    cli
    -0.63
     Messages
    -0.63
    calling
    -0.63
    ergy
    -0.61
    POSITIVE LOGITS
     endure
    1.09
     confront
    1.01
     reckon
    1.00
     rethink
    0.96
     fend
    0.93
     reconsider
    0.92
     flee
    0.87
     cancel
    0.87
     abandon
    0.86
     undertake
    0.84
    Act Density 0.068%

    No Known Activations