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

    Neuronpedia

    INDEX
    Explanations

    numbered list explanation

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

    formatted section headings and ordered list markers (numbers/letters/Roman numerals, often bolded or followed by a period) indicating structured subsections.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     tuple
    0.23
     cải
    0.23
     biais
    0.21
     intang
    0.21
     réduire
    0.21
     diminue
    0.21
     jiné
    0.21
     emotes
    0.21
     mêmes
    0.21
     erad
    0.21
    POSITIVE LOGITS
    Mga
    0.28
     Какие
    0.28
    first
    0.26
    Какие
    0.26
    Which
    0.25
    Detailed
    0.25
     본격
    0.25
    What
    0.24
    How
    0.24
    which
    0.24
    Act Density 1.613%

    No Known Activations