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

    Neuronpedia

    INDEX
    Explanations

    end-of-turn markers in a dialogue or discussion format

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

    mathematical and symbolic expression

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

    It detects the end-of-turn / segment boundary token (i.e., markers indicating the end of a turn).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    -0.36
    BeginContext
    -0.35
    Pautan
    -0.34
    RectangleBorder
    -0.33
    蚪
    -0.33
    ]<<
    -0.33
     архивлан
    -0.33
     venons
    -0.32
    Спасылкі
    -0.32
    apimachinery
    -0.32
    POSITIVE LOGITS
    didSet
    0.68
    enderror
    0.59
    UnusedPrivate
    0.58
    COUVER
    0.54
    𖧹
    0.53
    ſelben
    0.52
    perk
    0.51
    喫茶
    0.51
    𖡻
    0.50
     betweenstory
    0.49
    Act Density 0.115%

    No Known Activations