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

    Neuronpedia

    INDEX
    Explanations

    objects and locations within a room or household

    oai_token-act-pair · gpt-3.5-turbo

    descriptive elements related to setting and environment

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    umbn
    -0.66
    emonic
    -0.63
    ãĥ©ãĥ³
    -0.63
    ozyg
    -0.61
     Fatal
    -0.61
    è¦ļéĨĴ
    -0.60
    atri
    -0.60
    ewitness
    -0.60
    kus
    -0.59
     dstg
    -0.58
    POSITIVE LOGITS
     challeng
    0.76
     wip
    0.74
     tidy
    0.71
     attic
    0.70
     horizont
    0.70
     tray
    0.68
    robe
    0.68
    ascript
    0.66
     radiator
    0.66
    frame
    0.66
    Act Density 0.342%

    No Known Activations