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

    Neuronpedia

    INDEX
    Explanations

    scientific writing

    np_max-act · gemini-2.0-flash

    scientific and technical terminology related to control systems, particularly in the context of PID (Proportional, Integral, Derivative) controllers and their optimization methods.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The main thing this neuron does is detect specialized technical jargon—especially acronyms and multi-syllabic scientific or engineering terms.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Retrofit
    -0.07
     关
    -0.07
     lut
    -0.07
    けて
    -0.06
    (sign
    -0.06
    ��
    -0.06
    mediately
    -0.06
    YLE
    -0.06
     minced
    -0.06
    EMBER
    -0.06
    POSITIVE LOGITS
     benches
    0.07
     напря
    0.06
    심
    0.06
    .FileName
    0.06
     discs
    0.06
    abc
    0.06
     Crus
    0.06
     Україна
    0.06
    gien
    0.06
     mercury
    0.06
    Act Density 0.295%

    No Known Activations