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

    Neuronpedia

    INDEX
    Explanations

    conversation snippets

    np_max-act · gemini-2.0-flash

    situations involving relationship dynamics and emotional connections.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    This neuron detects speaker-identifying tokens or turn labels (e.g., “Z:”, “O:”, names) marking who’s speaking.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    CEPTION
    -0.07
     ise
    -0.06
     Px
    -0.06
    strap
    -0.06
     Ax
    -0.06
     okul
    -0.06
    Ax
    -0.06
    	System
    -0.06
     decoder
    -0.06
     footwear
    -0.06
    POSITIVE LOGITS
    іка
    0.07
    自分の
    0.07
     ấn
    0.06
     isOpen
    0.06
    віт
    0.06
    _CMD
    0.06
     سال
    0.06
    建议
    0.06
    /Create
    0.06
     bruk
    0.06
    Act Density 0.064%

    No Known Activations