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

    Neuronpedia

    INDEX
    Explanations

    words related to conducting trials or experiments

    oai_token-act-pair · gpt-3.5-turbo

    the repeated instances of the word "test" and its variations, indicating a focus on evaluation or experimentation

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    taboola
    -0.72
    brance
    -0.71
    lance
    -0.69
    nown
    -0.68
    \-
    -0.66
    vation
    -0.65
    ometown
    -0.65
    nick
    -0.63
    apo
    -0.63
    ffe
    -0.63
    POSITIVE LOGITS
     hypotheses
    0.98
    osterone
    0.88
    ifies
    0.81
    orously
    0.80
     whether
    0.80
     hypothesis
    0.70
    ify
    0.69
    imon
    0.69
     feasibility
    0.68
    onom
    0.68
    Act Density 0.078%

    No Known Activations