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

    Neuronpedia

    INDEX
    Explanations

    phrases related to train derailments

    oai_token-act-pair · gpt-3.5-turbo

    references to train derailments and related events

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Kinnikuman
    -0.79
    ternity
    -0.78
    isations
    -0.73
    archs
    -0.70
    ples
    -0.70
    isdom
    -0.68
    orneys
    -0.68
    estate
    -0.68
    hesda
    -0.66
    atar
    -0.65
    POSITIVE LOGITS
     derail
    1.46
     derailed
    1.11
    heimer
    0.84
    crew
    0.81
    wagon
    0.80
     Amtrak
    0.78
     Dickinson
    0.74
    ashtra
    0.70
     freight
    0.69
     rupture
    0.69
    Act Density 0.029%

    No Known Activations