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 fires on Makefile‐style variable definitions—especially uppercase build/path identifiers (e.g. BUILD, DIR, PATH) and their assignment tokens.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    makefile directives and configurations related to building software.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    _defined
    -0.07
    (fin
    -0.06
    U
    -0.06
    enedor
    -0.06
     참가
    -0.06
    (gl
    -0.06
     hầu
    -0.06
    され
    -0.06
    _Double
    -0.06
    isFunction
    -0.06
    POSITIVE LOGITS
     нас
    0.08
     lượt
    0.07
     nombreux
    0.06
     béné
    0.06
    砲
    0.06
    εβ
    0.06
     hangi
    0.06
     Drew
    0.06
     phổ
    0.06
     Nicholas
    0.06
    Act Density 0.008%

    No Known Activations