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

    Neuronpedia

    INDEX
    Explanations

    file directory structure

    np_max-act · gemini-2.0-flash

    This neuron detects the ASCII-art tree structure characters (e.g. “├──”, “└──”, folder/path separators) used to display a file/directory tree.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ト
    -0.07
     <=
    -0.07
     garnered
    -0.07
    ibe
    -0.07
    анных
    -0.06
    anki
    -0.06
    ertura
    -0.06
    냐
    -0.06
     millennium
    -0.06
    ietet
    -0.06
    POSITIVE LOGITS
     distract
    0.06
    _Build
    0.06
     شكل
    0.06
     Projekt
    0.06
     uplat
    0.06
     Bundy
    0.06
    _Tr
    0.06
     Voc
    0.06
    /MPL
    0.06
     rozvoj
    0.06
    Act Density 0.016%

    No Known Activations