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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    The neuron activates on programming language keywords and syntax tokens (e.g. import, var, function, new, etc.), effectively spotting snippets of source code.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    食べ
    -0.07
    available
    -0.06
    gender
    -0.06
     السعودية
    -0.06
     RTS
    -0.06
     fus
    -0.06
    руд
    -0.06
     ของ
    -0.06
    ωση
    -0.06
    _PROPERTY
    -0.06
    POSITIVE LOGITS
     trval
    0.07
     OPTION
    0.06
     Lego
    0.06
     tighter
    0.06
    -bodied
    0.06
    TexImage
    0.06
     xlabel
    0.06
    Avatar
    0.06
     conhe
    0.05
     nano
    0.05
    Act Density 0.024%

    No Known Activations