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

    Neuronpedia

    INDEX
    Explanations

    phrases related to controversial or potentially harmful topics, such as racially motivated attacks, compromised identity keys, health concerns linked to pesticides, and claims of false statements

    oai_token-act-pair · gpt-3.5-turbo

    references to social and legal issues, particularly those involving crime, politics, and public sentiment

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Darling
    -0.52
    Jr
    -0.52
     concess
    -0.49
    mbuds
    -0.47
     Kop
    -0.47
     overe
    -0.46
    retty
    -0.46
    sit
    -0.46
    educ
    -0.46
    advert
    -0.46
    POSITIVE LOGITS
    )?
    0.94
    ¶
    0.84
     Belfast
    0.79
     ):
    0.76
     constitutes
    0.75
     violates
    0.75
    .--
    0.75
    .–
    0.72
     Copyright
    0.72
    "?
    0.72
    Act Density 1.375%

    No Known Activations