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

    Neuronpedia

    INDEX
    Explanations

    phrases related to negative consequences or obstacles

    oai_token-act-pair · gpt-3.5-turbo

    references to economic impacts and risks associated with consumers and industries

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ³³³³
    -0.73
    eth
    -0.69
    ,[
    -0.66
    hest
    -0.65
    onomy
    -0.64
    edom
    -0.64
    ACTED
    -0.64
    kind
    -0.64
    wine
    -0.64
    ãģ®éŃĶ
    -0.63
    POSITIVE LOGITS
     unsuspecting
    1.19
     costly
    1.07
     otherwise
    1.05
     subsequent
    1.03
     pesky
    1.00
     unforeseen
    1.00
     newcomers
    1.00
     later
    0.95
     larger
    0.93
     other
    0.93
    Act Density 0.791%

    No Known Activations