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

    Neuronpedia

    INDEX
    Explanations

    multi-part phrases that build up to a concept or idea

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to specific organizations or clubs

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     channels
    -0.74
     lines
    -0.66
     embassies
    -0.65
     Sections
    -0.64
    rooms
    -0.64
     dots
    -0.62
     Guns
    -0.62
     Soldiers
    -0.62
     ports
    -0.61
     Units
    -0.61
    POSITIVE LOGITS
     akin
    0.91
     standpoint
    0.87
    iator
    0.84
     arrangement
    0.84
    esque
    0.82
    rower
    0.79
    ilege
    0.75
    otomy
    0.75
    ressor
    0.74
     converter
    0.74
    Act Density 0.603%

    No Known Activations