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

    Neuronpedia

    INDEX
    Explanations

    >

    np_max-act · gemini-2.0-flash

    This neuron detects occurrences of “of” in include paths, i.e. references to the Linux device-tree (“open firmware”) headers.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     rop
    -0.07
    allas
    -0.06
     Kavanaugh
    -0.06
     Shutterstock
    -0.06
    Ram
    -0.06
     Should
    -0.06
     Fil
    -0.06
    ISK
    -0.06
    883
    -0.06
    -cut
    -0.05
    POSITIVE LOGITS
    (material
    0.08
     solo
    0.07
     realtime
    0.07
    leşik
    0.07
    เ
    0.07
    zion
    0.06
     contentValues
    0.06
    _variance
    0.06
    чет
    0.06
     essentially
    0.06
    Act Density 0.005%

    No Known Activations