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

    Neuronpedia

    INDEX
    Explanations

    comma

    np_max-act · gemini-2.0-flash

    This neuron is tuned to detect the phrase “no matter” (particularly when used to open a concessive clause, often followed by a comma).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     lamps
    -0.08
    'eau
    -0.07
     är
    -0.07
    .mar
    -0.07
     INV
    -0.07
     STREAM
    -0.06
    ican
    -0.06
    LET
    -0.06
     bet
    -0.06
    니스
    -0.06
    POSITIVE LOGITS
    ]=>
    0.07
    cooldown
    0.06
     attribution
    0.06
     зави
    0.06
    TK
    0.06
     ideology
    0.06
     CSRF
    0.06
    Я
    0.06
     CLL
    0.05
    slideUp
    0.05
    Act Density 0.023%

    No Known Activations