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

    Neuronpedia

    INDEX
    Explanations

    rare word fragments

    np_max-act · gemini-2.0-flash

    The neuron fires on substrings of proper names—especially brand or company names—marking those distinctive capitalized chunks in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     nto
    -0.06
    prof
    -0.06
     steril
    -0.06
    variable
    -0.06
    !".
    -0.06
    privileged
    -0.06
    üsü
    -0.06
    *i
    -0.06
     stair
    -0.06
     Kant
    -0.06
    POSITIVE LOGITS
     bày
    0.07
     nullable
    0.07
    decess
    0.07
    ність
    0.06
     vỏ
    0.06
     بوده
    0.06
    .ToBoolean
    0.06
     waypoints
    0.06
    ниця
    0.06
    无
    0.06
    Act Density 0.105%

    No Known Activations