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

    Neuronpedia

    INDEX
    Explanations

    unexpected or surprising events and situations

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
     throats
    -0.84
     favored
    -0.80
     hemor
    -0.76
     throat
    -0.75
    ailability
    -0.75
     approved
    -0.74
    approved
    -0.74
    ona
    -0.73
    illes
    -0.72
    stood
    -0.70
    POSITIVE LOGITS
     Sharif
    0.92
     Gaw
    0.90
     juxtap
    0.89
     parallels
    0.86
     irony
    0.79
     how
    0.79
     EDITION
    0.78
     Sturgeon
    0.78
     Vaugh
    0.77
     Manitoba
    0.77
    Act Density 1.762%

    No Known Activations