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

    Neuronpedia

    INDEX
    Explanations

    start of commands or requests

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

    phrases indicating imperative user instructions or task requests, often at the start of a prompt and across multiple languages.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     tunt
    0.27
     audiovis
    0.26
     canales
    0.26
     einzelnen
    0.26
     fisik
    0.26
     Artis
    0.26
     flotte
    0.26
     sienten
    0.26
     व्याव
    0.26
     അവർ
    0.26
    POSITIVE LOGITS
    様専用
    0.29
    または
    0.28
     morphism
    0.28
    数组
    0.27
    ԁ
    0.27
     algebraically
    0.27
     outperforms
    0.27
    oq
    0.26
    oeste
    0.25
    定義
    0.25
    Act Density 0.401%

    No Known Activations