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

    Neuronpedia

    INDEX
    Explanations

    phrases related to considerations or aspects of a specific topic

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    orr
    -0.65
    stall
    -0.64
     crawl
    -0.62
    CHA
    -0.60
     cascade
    -0.59
     tangled
    -0.59
    tick
    -0.59
     cycle
    -0.59
    right
    -0.57
    gain
    -0.56
    POSITIVE LOGITS
     regards
    4.36
     regard
    2.38
     respects
    1.86
     respect
    1.45
     relation
    1.15
     terms
    1.14
    Regarding
    1.12
     hopes
    1.07
     apologies
    0.99
     endeavors
    0.98
    Act Density 0.005%

    No Known Activations