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

    Neuronpedia

    INDEX
    Explanations

    phrases related to violence against civilians

    oai_token-act-pair · gpt-3.5-turbo

    mentions of civilians in contexts of conflict or war

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ologies
    -0.88
    href
    -0.78
    ology
    -0.72
    Vert
    -0.70
    Smith
    -0.69
    Stud
    -0.69
    ENCE
    -0.68
    SourceFile
    -0.68
    TOP
    -0.67
    Magikarp
    -0.67
    POSITIVE LOGITS
     casualties
    0.86
     fleeing
    0.85
    folk
    0.83
     combatants
    0.79
     slain
    0.78
     bystanders
    0.78
     displaced
    0.78
     hostages
    0.76
     harmed
    0.74
     massac
    0.74
    Act Density 0.019%

    No Known Activations