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

    Neuronpedia

    INDEX
    Explanations

    words related to suppressing or quelling actions or emotions

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    anuts
    -0.84
    inav
    -0.77
    ammy
    -0.75
    abet
    -0.75
    olds
    -0.74
    olf
    -0.71
    better
    -0.71
    eday
    -0.70
    optional
    -0.70
    nih
    -0.69
    POSITIVE LOGITS
     dissent
    0.97
     distractions
    0.88
     impulses
    0.84
     hostilities
    0.84
     havoc
    0.83
     pests
    0.83
     dissenting
    0.82
     weeds
    0.81
     temptation
    0.80
     intr
    0.80
    Act Density 0.180%

    No Known Activations