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

    Neuronpedia

    INDEX
    Explanations

    situations involving violence or crime

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    DAQ
    -0.86
    notation
    -0.85
    ãĤ´ãĥ³
    -0.83
    tions
    -0.82
    utical
    -0.81
    Works
    -0.80
    pring
    -0.77
    iture
    -0.77
    Whe
    -0.77
    inventoryQuantity
    -0.76
    POSITIVE LOGITS
     gunmen
    1.17
     thugs
    1.12
     robbers
    1.12
     unidentified
    1.09
     assailants
    1.04
     bandits
    0.99
     thieves
    0.99
     gunfire
    0.97
     masked
    0.96
     passers
    0.94
    Act Density 0.175%

    No Known Activations