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

    Neuronpedia

    INDEX
    Explanations

    phrases related to maintaining, preserving, or holding things together

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to maintaining stability and well-being

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     fit
    -0.65
    haps
    -0.64
    FER
    -0.59
    ouf
    -0.58
     merge
    -0.58
     lump
    -0.58
     fitting
    -0.58
     converter
    -0.58
     descending
    -0.57
    fits
    -0.57
    POSITIVE LOGITS
     indefinitely
    1.06
     until
    0.96
     whilst
    0.89
     till
    0.84
     forever
    0.82
     throughout
    0.82
     lest
    0.81
     despite
    0.81
     amid
    0.80
     uninterrupted
    0.79
    Act Density 0.177%

    No Known Activations