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

    Neuronpedia

    INDEX
    Explanations

    the phrase "get rid of" followed by a noun or verb

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating the concept of elimination or removal

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Cosponsors
    -0.68
     confir
    -0.66
    Rate
    -0.63
    INFO
    -0.61
    alian
    -0.61
    STATE
    -0.60
     challeng
    -0.60
    eller
    -0.60
    MODE
    -0.58
    Act
    -0.57
    POSITIVE LOGITS
     carc
    0.76
     unwanted
    0.76
     pesky
    0.75
     stray
    0.74
     EVs
    0.72
     anything
    0.69
    olphins
    0.69
     any
    0.67
     virginity
    0.65
     loopholes
    0.65
    Act Density 0.041%

    No Known Activations