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

    Neuronpedia

    INDEX
    Explanations

    phrases related to justification or reasoning

    oai_token-act-pair · gpt-3.5-turbo

    phrases that contain the terms "basis" and "grounds" which are often used to indicate reasoning or justification

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ascus
    -0.79
    icer
    -0.74
    quer
    -0.71
    Town
    -0.68
    oppable
    -0.66
    ilde
    -0.65
    edo
    -0.65
    agra
    -0.65
     Sparrow
    -0.64
    resent
    -0.64
    POSITIVE LOGITS
     basis
    1.03
     grounds
    1.03
     Grounds
    0.90
     premise
    0.84
    tyard
    0.78
    grounds
    0.77
    fy
    0.74
    ifiers
    0.73
    mination
    0.73
     principle
    0.73
    Act Density 0.032%

    No Known Activations