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

    Neuronpedia

    INDEX
    Explanations

    the word "how" and its variations, indicating inquiries about processes or methods

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     raft
    -0.86
     Schulte
    -0.85
     Gegenteil
    -0.82
     Milne
    -0.80
     Münch
    -0.76
     tdi
    -0.76
     occaf
    -0.75
     Cæsar
    -0.75
    گي
    -0.75
     Balik
    -0.75
    POSITIVE LOGITS
     how
    1.88
     How
    1.57
     HOW
    1.57
    How
    1.46
    HOW
    1.37
    how
    1.25
     Howell
    1.07
    howto
    1.00
     Howe
    0.99
     what
    0.99
    Act Density 0.074%

    No Known Activations