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

    Neuronpedia

    INDEX
    Explanations

    file extensions

    np_max-act · gemini-2.0-flash

    The neuron fires on tokens that look like file names or file paths with extensions (e.g. “foo.txt”, “bar.cpp”, “baz.yaml”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    IAS
    -0.07
    .shiro
    -0.07
    -so
    -0.06
    uffer
    -0.06
    mpjes
    -0.06
    brew
    -0.06
    -php
    -0.06
     Contents
    -0.06
     toddlers
    -0.06
    -five
    -0.06
    POSITIVE LOGITS
     بخ
    0.06
     míst
    0.06
     kli
    0.06
     �
    0.06
     Nu
    0.06
     gunfire
    0.06
    оваться
    0.06
    actal
    0.06
    ("/{
    0.06
    _art
    0.06
    Act Density 0.014%

    No Known Activations