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

    Neuronpedia

    INDEX
    Explanations

    phrases or words related to comparisons

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "against"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    hops
    -0.91
    overed
    -0.81
     liner
    -0.77
    aird
    -0.76
    operated
    -0.76
    DragonMagazine
    -0.75
    jri
    -0.74
    lus
    -0.72
    details
    -0.71
    assy
    -0.71
    POSITIVE LOGITS
     whom
    0.83
     anybody
    0.80
     against
    0.77
     them
    0.69
     somebody
    0.67
     behalf
    0.67
     anyone
    0.66
     Humanity
    0.65
     adversity
    0.65
     Against
    0.65
    Act Density 0.046%

    No Known Activations