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

    Neuronpedia

    INDEX
    Explanations

    phrases related to watching or observing, especially in intense or critical situations

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Negative Logits
     konkre
    -0.62
     benzin
    -0.61
     hek
    -0.59
     treff
    -0.57
     kafe
    -0.57
     kompres
    -0.56
     erk
    -0.56
     garan
    -0.56
     vermel
    -0.55
     ekster
    -0.55
    POSITIVE LOGITS
     watching
    0.74
     watched
    0.71
    watched
    0.70
    Watched
    0.68
     watch
    0.68
     watches
    0.67
    watching
    0.66
     Watched
    0.65
     contextLoads
    0.61
     Watching
    0.61
    Act Density 0.188%

    No Known Activations