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

    Neuronpedia

    INDEX
    Explanations

    the word "whenever" and its variations, indicating a focus on conditional or temporal contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Samoa
    -0.71
     Daly
    -0.67
     sich
    -0.66
    ugat
    -0.63
    obs
    -0.63
    ynh
    -0.62
     crazy
    -0.61
     paese
    -0.59
    Already
    -0.58
    cách
    -0.58
    POSITIVE LOGITS
     whenever
    2.34
    whenever
    2.30
     Whenever
    2.21
    Whenever
    2.09
     wherever
    1.24
     Wherever
    1.18
     everytime
    1.05
    Wherever
    1.04
     Anytime
    1.01
     anytime
    0.99
    Act Density 0.127%

    No Known Activations