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

    Neuronpedia

    INDEX
    Explanations

    words related to different levels or features of floors

    oai_token-act-pair · gpt-3.5-turbo

    references to "floors" and "ceilings" in buildings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    vous
    -0.85
    FUL
    -0.70
     CNS
    -0.69
    nia
    -0.67
    wik
    -0.65
     Canad
    -0.64
     Blessed
    -0.63
    enegger
    -0.63
    nell
    -0.61
    language
    -0.61
    POSITIVE LOGITS
     floors
    1.30
    floor
    0.93
     ceilings
    0.93
    raper
    0.90
     carp
    0.90
     Floor
    0.89
     attic
    0.86
     roofs
    0.84
    qus
    0.84
     floor
    0.82
    Act Density 0.010%

    No Known Activations