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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating possibility

    oai_token-act-pair · gpt-3.5-turbo

    phrases expressing possibilities or conjectures

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ciating
    -0.90
    utch
    -0.73
    oba
    -0.69
    ggles
    -0.67
    hesion
    -0.67
    ving
    -0.66
    ophe
    -0.66
    pes
    -0.65
    equ
    -0.65
    usalem
    -0.65
    POSITIVE LOGITS
     underest
    0.77
     they
    0.77
     someday
    0.72
     exaggeration
    0.69
     coincidence
    0.67
     premature
    0.67
     underestimate
    0.66
     that
    0.65
     there
    0.64
     Rasm
    0.63
    Act Density 0.094%

    No Known Activations