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

    Neuronpedia

    INDEX
    Explanations

    instances of violence or tragedy

    oai_token-act-pair · gpt-3.5-turbo

    instances of violent events or casualties

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    pure
    -0.72
    ãĤ¦ãĤ¹
    -0.70
    medium
    -0.69
    mosp
    -0.69
     comprehension
    -0.68
    simple
    -0.64
    Medium
    -0.64
     somew
    -0.63
    6000
    -0.63
     imagination
    -0.63
    POSITIVE LOGITS
     murdered
    1.26
     assassinated
    1.24
     slain
    1.19
     ousted
    1.14
     killed
    1.13
     gunned
    1.10
     massac
    1.08
     toppled
    1.06
     disappeared
    1.05
     died
    1.04
    Act Density 0.906%

    No Known Activations