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

    Neuronpedia

    INDEX
    Explanations

    Document beginning with a date

    np_max-act · gemini-2.0-flash

    The neuron detects the comma (and immediately following space) used in formatted dates like “Feb 17, 2017” (i.e. the separator between day and year).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    掌
    -0.07
     及
    -0.07
    oned
    -0.07
     pie
    -0.07
     Flags
    -0.06
     Mısır
    -0.06
    “我
    -0.06
     Pie
    -0.06
    :S
    -0.06
    "]["
    -0.06
    POSITIVE LOGITS
    documentation
    0.06
    viron
    0.06
     harmon
    0.06
     التن
    0.06
    .pem
    0.06
    	scale
    0.06
    ились
    0.06
     anche
    0.06
     Denis
    0.06
    .jav
    0.06
    Act Density 0.005%

    No Known Activations