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

    Neuronpedia

    INDEX
    Explanations

    different forms of the word "upgrade" or "downgrade"

    oai_token-act-pair · gpt-3.5-turbo

    terms related to rating changes, particularly downgrades and upgrades

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    stack
    -0.88
    advertisement
    -0.87
     Dover
    -0.71
    vous
    -0.71
     Caption
    -0.67
     Acts
    -0.67
    Pa
    -0.66
     Cthulhu
    -0.65
    eye
    -0.64
    laws
    -0.64
    POSITIVE LOGITS
    graded
    1.50
     downgrade
    1.31
    grading
    1.26
     upgraded
    1.13
    grades
    1.05
    xual
    1.02
     upgrades
    0.98
     upgrading
    0.91
     upgrade
    0.90
     grades
    0.88
    Act Density 0.006%

    No Known Activations