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

    Neuronpedia

    INDEX
    Explanations

    words related to justification or reasoning for actions

    oai_token-act-pair · gpt-3.5-turbo

    terms related to justification and reasoning

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ummer
    -0.84
     Carbuncle
    -0.76
    chron
    -0.72
    ammy
    -0.72
    Hop
    -0.71
    estone
    -0.70
    berry
    -0.69
    ept
    -0.68
    grass
    -0.68
    thritis
    -0.67
    POSITIVE LOGITS
     justifying
    1.00
     justification
    0.89
     justify
    0.87
     justifies
    0.87
     why
    0.79
    ably
    0.74
     justified
    0.73
     excuses
    0.72
     guiActiveUn
    0.70
    ighed
    0.69
    Act Density 0.032%

    No Known Activations