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

    Neuronpedia

    INDEX
    Explanations

    30, 0

    np_max-act · gemini-2.0-flash

    This neuron activates on hexadecimal escape or percent-encoding sequences (e.g., “\uXXXX” or “%XX”) in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Stage
    -0.06
    الى
    -0.06
    Day
    -0.06
    安全
    -0.06
    paragraph
    -0.06
     Lim
    -0.06
    oyo
    -0.06
    Sun
    -0.06
    eng
    -0.06
    できない
    -0.06
    POSITIVE LOGITS
     Covered
    0.07
    -TV
    0.07
    .)↵↵
    0.07
    [Any
    0.06
    (('
    0.06
    ellite
    0.06
    (DialogInterface
    0.06
     APIs
    0.06
     RAID
    0.06
    /MPL
    0.06
    Act Density 0.005%

    No Known Activations