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

    Neuronpedia

    INDEX
    Explanations

    accompanying items or information related to a main subject

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "accompanying" and variations, indicating a focus on content that is supplementary or supportive

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    $$
    -0.73
    Start
    -0.68
    wer
    -0.65
     ***
    -0.63
    Token
    -0.58
    **
    -0.57
    once
    -0.57
    iable
    -0.57
    rals
    -0.57
     Quit
    -0.56
    POSITIVE LOGITS
     accompanying
    3.90
     accompany
    2.29
     accompanies
    1.98
     attendant
    1.57
     accompan
    1.39
     accompanied
    1.38
     ensuing
    1.35
    accompan
    1.31
     companion
    1.22
     underlying
    1.07
    Act Density 0.012%

    No Known Activations