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

    Neuronpedia

    INDEX
    Explanations

    question marks and periods

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

    punctuation that marks sentence boundaries and queries, especially question marks (and often periods/commas).

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     stri
    0.32
    FindFileData
    0.32
     formu
    0.31
    gie
    0.30
    ड़
    0.30
     algorithmic
    0.30
     restric
    0.29
     probabilities
    0.29
     grotes
    0.29
     dissol
    0.29
    POSITIVE LOGITS
     क्योंकि
    0.41
     It
    0.41
     কেননা
    0.40
     You
    0.40
     Öncelikle
    0.38
     I
    0.38
     nhưng
    0.37
    html
    0.37
     Because
    0.37
     প্রেসিডেন্ট
    0.37
    Act Density 0.579%

    No Known Activations