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

    Neuronpedia

    INDEX
    Explanations

    phrases expressing personal thoughts or reflections

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
     Publication
    -0.66
    Site
    -0.65
    ullivan
    -0.65
    Publisher
    -0.63
     Subcommittee
    -0.63
     Spons
    -0.63
     NYT
    -0.62
     CLR
    -0.62
     Hole
    -0.61
     reviewer
    -0.60
    POSITIVE LOGITS
     gotta
    1.09
     wanna
    1.02
    'm
    0.99
     feel
    0.98
     got
    0.91
     think
    0.90
     grav
    0.90
     get
    0.88
     remember
    0.88
    've
    0.86
    Act Density 0.348%

    No Known Activations