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

    Neuronpedia

    INDEX
    Explanations

    embellish adorned

    np_max-act · gemini-2.0-flash

    This neuron lights up on the model’s internal control or markup tokens (e.g. header/metadata tags) rather than on normal content words.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    预览
    -0.06
    Allocator
    -0.06
    zzle
    -0.06
    _PUS
    -0.06
    worksheet
    -0.06
    atology
    -0.06
    Cond
    -0.06
    -links
    -0.06
    UDIO
    -0.06
     Janeiro
    -0.06
    POSITIVE LOGITS
     which
    0.07
     Feather
    0.07
     periodo
    0.07
     Newtonsoft
    0.07
     adorned
    0.07
     attire
    0.07
     
    0.07
     одно
    0.07
     зміни
    0.06
    >Create
    0.06
    Act Density 0.012%

    No Known Activations