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

    Neuronpedia

    INDEX
    Explanations

    conjunctions indicating contrasting or opposing ideas

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "But" indicating a contrast or shift in the narrative

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    zona
    -0.72
    cellent
    -0.69
    Individual
    -0.69
    etermin
    -0.68
    keyes
    -0.67
    acceptable
    -0.67
    arest
    -0.67
    irms
    -0.64
    road
    -0.63
    industrial
    -0.63
    POSITIVE LOGITS
     shortly
    1.18
     alas
    1.17
     soon
    1.14
     afterward
    1.02
     luckily
    1.01
     after
    1.01
     amidst
    0.99
     despite
    0.99
     amid
    0.98
     unlike
    0.98
    Act Density 0.248%

    No Known Activations