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

    Neuronpedia

    INDEX
    Explanations

    the word "merely" in the text

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "merely" and its variants to indicate trivialization or minimizing statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     hi
    -0.61
    rain
    -0.57
     Awesome
    -0.57
     Calls
    -0.56
    Di
    -0.56
    SI
    -0.56
    ez
    -0.56
    aura
    -0.56
    oken
    -0.55
    Hi
    -0.55
    POSITIVE LOGITS
     merely
    3.23
     simply
    2.16
     mere
    1.99
     purely
    1.68
     solely
    1.59
    mere
    1.55
     strictly
    1.18
     just
    1.18
     only
    1.15
    Simply
    1.14
    Act Density 0.007%

    No Known Activations