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

    Neuronpedia

    INDEX
    Explanations

    phrases related to ensuring things are done correctly or in a specified manner

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to ensuring safety, compliance, and proper functioning in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     pity
    -0.73
     unlucky
    -0.67
     glim
    -0.65
    famous
    -0.63
     unwitting
    -0.63
     bluff
    -0.62
     Worse
    -0.61
     wonder
    -0.59
     whining
    -0.59
     delusional
    -0.58
    POSITIVE LOGITS
     irrespective
    1.02
     throughout
    1.01
     regardless
    1.00
     wherever
    0.91
     consistent
    0.85
     whenever
    0.85
     across
    0.79
     whilst
    0.78
    cellence
    0.76
     accordingly
    0.76
    Act Density 0.458%

    No Known Activations