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

    Neuronpedia

    INDEX
    Explanations

    phrases conveying disappointment or dissatisfaction

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    abase
    -0.61
     UL
    -0.60
    asio
    -0.60
     Basics
    -0.58
    amental
    -0.58
    INC
    -0.57
     attest
    -0.57
     unforgettable
    -0.55
    ieu
    -0.55
     beware
    -0.55
    POSITIVE LOGITS
     suddenly
    0.98
     bothered
    0.89
     chose
    0.88
     resorted
    0.87
     dared
    0.86
     survived
    0.86
     didnt
    0.85
     somehow
    0.84
     ever
    0.84
     finally
    0.83
    Act Density 0.295%

    No Known Activations