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

    Neuronpedia

    INDEX
    Explanations

    instances where references are made to being transparent about mistakes, including grammar-related errors and personal errors

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Negative Logits
     Majest
    -0.74
     vnt
    -0.70
     addirittura
    -0.66
     impra
    -0.66
     Juf
    -0.65
     fta
    -0.65
     Augu
    -0.64
     thut
    -0.64
     délib
    -0.63
     Bibl
    -0.62
    POSITIVE LOGITS
    <bos>
    0.80
     inevitable
    0.65
     inevitably
    0.65
     sometime
    0.60
     occasional
    0.59
     occasionally
    0.56
     eventually
    0.56
    SneakyThrows
    0.54
    CodedInputStream
    0.53
     somewhere
    0.50
    Act Density 0.367%

    No Known Activations