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

    Neuronpedia

    INDEX
    Explanations

    numeric amounts and goals expressed in various forms of text

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to achieving goals or milestones

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     sake
    -0.67
     followed
    -0.66
     preceded
    -0.65
    sbm
    -0.64
     alone
    -0.63
     arose
    -0.63
     Tud
    -0.62
     endured
    -0.60
     mattered
    -0.58
     transpired
    -0.58
    POSITIVE LOGITS
     heights
    1.14
     milestone
    1.04
     conclusion
    1.00
     threshold
    0.98
     halfway
    0.94
    reshold
    0.92
     plateau
    0.90
     pinnacle
    0.90
     milestones
    0.87
    levels
    0.86
    Act Density 0.149%

    No Known Activations