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

    Neuronpedia

    INDEX
    Explanations

    initial release

    np_max-act · gemini-2.0-flash

    The neuron is picking out version identifiers and release headings (e.g. numeric version numbers and “Initial release” lines) in changelog texts.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     treated
    -0.07
     pounded
    -0.06
    ailing
    -0.06
    ++;↵↵
    -0.06
    _answer
    -0.06
     vaping
    -0.06
    TypeName
    -0.06
    Ign
    -0.06
     SUPPORT
    -0.06
     iPhone
    -0.06
    POSITIVE LOGITS
     previously
    0.07
     initial
    0.07
    ‬↵
    0.07
     рев
    0.07
    lbrakk
    0.07
     faaliyet
    0.06
     شه
    0.06
     우리
    0.06
    (PDO
    0.06
    formerly
    0.06
    Act Density 0.002%

    No Known Activations