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

    Neuronpedia

    INDEX
    Explanations

    line

    np_max-act · gemini-2.0-flash

    The neuron detects mentions of “command-line arguments” (i.e., the hyphenated term “command-line” in the context of listing arguments).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ources
    -0.07
     HashSet
    -0.07
    .NonNull
    -0.06
     ایران
    -0.06
    智能
    -0.06
     Jacqu
    -0.06
    anguages
    -0.06
    	fun
    -0.06
    	code
    -0.06
     andere
    -0.06
    POSITIVE LOGITS
    elin
    0.07
     driveway
    0.07
    larg
    0.06
    lit
    0.06
    diag
    0.06
    lene
    0.06
    ruh
    0.06
     neden
    0.06
     지나
    0.06
     kla
    0.06
    Act Density 0.004%

    No Known Activations