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

    Neuronpedia

    INDEX
    Explanations

    instances of the apostrophe in text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    foreign words and archaic spellings

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron detects the apostrophe character (the single-quote) typically used in contractions.

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    ↵↵
    -0.52
    ↵
    -0.47
     and
    -0.44
     (
    -0.43
     The
    -0.43
     
    -0.43
     etc
    -0.42
    and
    -0.38
      
    -0.38
    as
    -0.37
    POSITIVE LOGITS
     zwiſchen
    1.29
    ſicht
    1.27
     imagui
    1.27
     ddelwed
    1.26
     Geiſt
    1.23
    ðsíða
    1.23
     queſta
    1.23
    ésultats
    1.23
    ロウィン
    1.22
     unſer
    1.22
    Act Density 2.269%

    No Known Activations