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

    Neuronpedia

    INDEX
    Explanations

    word pairs with some preceding ones' last letters connecting to the next ones' first letters

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to future plans and aspirations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Attacks
    -0.81
     advers
    -0.73
     Corruption
    -0.72
     analys
    -0.69
    Overview
    -0.68
     Overview
    -0.68
     CIS
    -0.68
     Analyst
    -0.67
     inaccur
    -0.67
     misrepresent
    -0.66
    POSITIVE LOGITS
     tidy
    1.05
     happily
    1.04
     snug
    0.96
     gladly
    0.95
     handy
    0.94
     snacks
    0.93
     tucked
    0.93
     comforting
    0.93
     thankful
    0.92
     spared
    0.92
    Act Density 0.792%

    No Known Activations