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

    Neuronpedia

    INDEX
    Explanations

    fiction excerpts

    np_max-act · gemini-2.0-flash

    This neuron detects special “speaker” or control tokens that mark system/user/assistant headers and in-role instructions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    узы
    -0.07
    %m
    -0.07
    ESSAGES
    -0.06
     assertions
    -0.06
    .getcwd
    -0.06
    InputStream
    -0.06
     subjected
    -0.06
    chain
    -0.06
     one
    -0.06
    GORITH
    -0.06
    POSITIVE LOGITS
     grind
    0.08
     tahun
    0.07
     govern
    0.07
     �
    0.07
    _display
    0.07
     exhibiting
    0.07
    ผม
    0.06
    ص
    0.06
    τικής
    0.06
     bát
    0.06
    Act Density 0.032%

    No Known Activations