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

    Neuronpedia

    INDEX
    Explanations

    my knowledge comes from data

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

    This neuron responds to asterisks used for emphasis or list/markdown formatting.

    oai_token-act-pair · o4-miniTriggered by @yooniel31

    self-referential passages where an AI model describes its identity, training/architecture, capabilities, and limitations.

    oai_token-act-pair · gpt-5Triggered by @yooniel31
    New Auto-Interp
    Negative Logits
    我們先
    0.49
    我們要
    0.47
     عايزين
    0.46
     આપણે
    0.46
     pihaknya
    0.46
     hassles
    0.43
     flimsy
    0.43
     menjual
    0.43
     buying
    0.41
    淔
    0.41
    POSITIVE LOGITS
     моих
    1.28
     моей
    1.23
     myself
    1.21
     мои
    1.20
     मेरे
    1.18
    我的
    1.16
     माझ्या
    1.10
     моего
    1.10
     আমার
    1.09
     meus
    1.09
    Act Density 0.152%

    No Known Activations