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

    Neuronpedia

    INDEX
    Explanations

    properties, features

    np_max-act · gemini-2.0-flash

    This neuron activates on adjectives and adverbs that describe performance metrics or favorable qualities (e.g. wide, excellent, fast, high, low, broad, superior).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    وسی
    -0.07
     XC
    -0.07
     phon
    -0.07
     maintained
    -0.07
     Сем
    -0.06
    _support
    -0.06
     mh
    -0.06
     Regex
    -0.06
    اری
    -0.06
    !:
    -0.06
    POSITIVE LOGITS
     ersten
    0.06
    .Navigator
    0.06
    Brightness
    0.06
    彼女
    0.06
     später
    0.06
    .atomic
    0.06
     后
    0.06
    cube
    0.06
     společně
    0.06
    .He
    0.06
    Act Density 0.292%

    No Known Activations