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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating contradiction or contrast

    oai_token-act-pair · gpt-3.5-turbo

    the word "despite" and its variations, indicating a focus on contrasting situations or conditions

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ecycle
    -0.87
    icter
    -0.86
    vantage
    -0.83
    enter
    -0.78
    apon
    -0.76
    oided
    -0.76
    endar
    -0.72
    iac
    -0.71
    farious
    -0.71
    eport
    -0.71
    POSITIVE LOGITS
     having
    1.20
     assurances
    1.13
     being
    1.08
     warnings
    1.03
     knowing
    1.03
     seeming
    1.02
     repeated
    1.02
     acknowledging
    1.02
     boasting
    1.01
     appearances
    0.99
    Act Density 0.039%

    No Known Activations