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

    Neuronpedia

    INDEX
    Explanations

    instances of the word "were" indicating past actions or states

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     liber
    -0.50
    lat
    -0.49
    lán
    -0.48
    azos
    -0.46
    ಞ
    -0.46
     fatt
    -0.45
    casecmp
    -0.45
     Sog
    -0.44
    ksin
    -0.44
    laps
    -0.44
    POSITIVE LOGITS
     were
    3.37
    Were
    2.95
     Were
    2.92
    were
    2.87
     WERE
    2.74
     weren
    2.22
    weren
    1.74
     étaient
    1.64
     voltak
    1.61
     были
    1.59
    Act Density 0.173%

    No Known Activations