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

    Neuronpedia

    INDEX
    Explanations

    negative or critical language used in a political context

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    emis
    -0.21
    gypt
    -0.21
    roo
    -0.20
    ":[
    -0.19
    hops
    -0.19
    ringe
    -0.18
     Sharing
    -0.18
    gnu
    -0.18
     transporting
    -0.17
    slave
    -0.17
    POSITIVE LOGITS
     revelations
    0.27
     rebuke
    0.26
    ly
    0.25
     blow
    0.24
     indictment
    0.23
     revelation
    0.23
     condemnation
    0.23
     accusation
    0.22
     headlines
    0.22
     accusations
    0.22
    Act Density 11.208%

    No Known Activations