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

    Neuronpedia

    INDEX
    Explanations

    positive adjectives

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "really" used as an intensifier

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    antry
    -0.91
    heid
    -0.84
    ansas
    -0.81
    icipated
    -0.70
    illary
    -0.69
    wright
    -0.69
    enary
    -0.68
    adelphia
    -0.68
    arthed
    -0.68
     Receiver
    -0.66
    POSITIVE LOGITS
     darn
    0.77
    FTWARE
    0.77
    ignment
    0.75
     appreciated
    0.74
     differentiated
    0.73
     liked
    0.72
     pissed
    0.71
     helpful
    0.69
     freaking
    0.69
     fucking
    0.69
    Act Density 0.054%

    No Known Activations