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

    Neuronpedia

    INDEX
    Explanations

    controversial or contrasting statements

    oai_token-act-pair · gpt-3.5-turbo

    references to generalizations and exceptions about different groups or categories of people

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     GENERAL
    -0.71
     Emin
    -0.69
    Motion
    -0.61
    oided
    -0.60
     Seat
    -0.60
     respectively
    -0.60
     Conservation
    -0.59
     NOTICE
    -0.58
     Consider
    -0.58
    ATURE
    -0.58
    POSITIVE LOGITS
     anymore
    1.68
     nor
    1.36
     necessarily
    0.95
    yet
    0.85
     anything
    0.82
     yet
    0.81
     anybody
    0.78
     anywhere
    0.78
    acea
    0.75
     slightest
    0.73
    Act Density 0.721%

    No Known Activations