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

    Neuronpedia

    INDEX
    Explanations

    phrases where someone is asserting or emphasizing a statement

    oai_token-act-pair · gpt-3.5-turbo

    phrases emphasizing belief or trust in statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    fax
    -0.72
    backer
    -0.69
    ific
    -0.65
    quit
    -0.65
    fix
    -0.64
    ãĤĵ
    -0.64
    hyde
    -0.63
    bats
    -0.62
     mail
    -0.61
     cort
    -0.60
    POSITIVE LOGITS
    uala
    0.74
    eele
    0.73
     independence
    0.71
     integrity
    0.71
    ournal
    0.70
     motives
    0.67
     Religion
    0.67
     sincerity
    0.67
     Citiz
    0.66
     authenticity
    0.65
    Act Density 0.160%

    No Known Activations