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

    Neuronpedia

    INDEX
    Explanations

    words related to significant emotional distress or suffering

    oai_token-act-pair · gpt-3.5-turbo

    expressions related to significant emotional distress or destruction

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    rouse
    -0.82
    chie
    -0.78
    perature
    -0.78
    abet
    -0.78
    akes
    -0.76
    rams
    -0.73
    outed
    -0.72
    asks
    -0.71
    reason
    -0.71
    ön
    -0.71
    POSITIVE LOGITS
     devastated
    0.98
    stal
    0.83
     traumat
    0.81
     adolesc
    0.77
     aback
    0.76
     Rican
    0.73
     Trident
    0.72
    DAQ
    0.69
     meltdown
    0.66
     stricken
    0.65
    Act Density 0.010%

    No Known Activations