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

    Neuronpedia

    INDEX
    Explanations

    phrases related to contradictory situations or contrasts

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Negative Logits
    isa
    -0.80
    vantage
    -0.79
    enter
    -0.78
    oided
    -0.77
    aird
    -0.77
    ahime
    -0.76
    ecycle
    -0.75
    venture
    -0.74
    seed
    -0.73
    asus
    -0.73
    POSITIVE LOGITS
     having
    1.05
     setbacks
    0.98
     appearances
    0.98
     assurances
    0.97
     being
    0.96
     warnings
    0.95
     objections
    0.94
     seeming
    0.93
     acknowledging
    0.90
     admitting
    0.90
    Act Density 0.077%

    No Known Activations