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

    Neuronpedia

    INDEX
    Explanations

    phrases related to being under the influence of something, often alcohol

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    ern
    -0.72
    boxing
    -0.67
     supra
    -0.66
    erness
    -0.66
    hops
    -0.65
    aries
    -0.65
    phabet
    -0.65
    aza
    -0.64
     ahead
    -0.64
    pring
    -0.63
    POSITIVE LOGITS
     guise
    1.26
     supervision
    1.21
     ausp
    1.18
     influence
    1.15
     microscope
    1.09
     radar
    1.09
     umbrella
    1.03
     guidance
    1.03
     pretext
    0.97
     hood
    0.96
    Act Density 0.051%

    No Known Activations