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

    Neuronpedia

    INDEX
    Explanations

    the presence of structured elements, such as brackets or mathematical notations

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

    numbers and code

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

    The neuron fires on numerical and math-related tokens (numbers, math-mode markers like "mathbb", "begin", and standalone numerals).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
     reciente
    -0.38
     récemment
    -0.37
     recent
    -0.37
     récentes
    -0.35
     yet
    -0.35
     recently
    -0.35
     récents
    -0.34
     recientemente
    -0.34
     soon
    -0.33
     bubble
    -0.33
    POSITIVE LOGITS
     CreateTagHelper
    0.83
    verwijspagina
    0.75
    ConstraintMaker
    0.73
    IsContent
    0.71
    Datuak
    0.70
     صوتيه
    0.68
    IntoConstraints
    0.64
    uxxxx
    0.64
     Normdatei
    0.63
     betweenstory
    0.61
    Act Density 0.242%

    No Known Activations