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

    Neuronpedia

    INDEX
    Explanations

    rounding numbers

    np_max-act · gemini-2.0-flash

    The neuron activates on runs of digits—i.e. numeric tokens (multi‐digit numbers or decimal‐fraction pieces).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    �
    -0.07
    raud
    -0.07
    هور
    -0.06
    .JFrame
    -0.06
    .Shapes
    -0.06
    -j
    -0.06
     ему
    -0.06
     Spears
    -0.06
    ุข
    -0.06
     wives
    -0.06
    POSITIVE LOGITS
    IGHL
    0.07
     участи
    0.06
    NF
    0.06
     menjadi
    0.06
     sitesi
    0.06
    itle
    0.06
    olation
    0.06
    (PARAM
    0.06
     Titanium
    0.06
     Miracle
    0.06
    Act Density 0.005%

    No Known Activations