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

    Neuronpedia

    INDEX
    Explanations

    's

    np_max-act · gemini-2.0-flash

    The neuron primarily activates on the Unicode replacement character “�,” i.e. unrecognized or garbled tokens.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    qml
    -0.07
     crushing
    -0.07
    -cluster
    -0.06
    rai
    -0.06
    utron
    -0.06
    ści
    -0.06
     THEIR
    -0.06
    Making
    -0.06
     CLEAN
    -0.06
    Craft
    -0.06
    POSITIVE LOGITS
     โ
    0.07
     واح
    0.07
    로서
    0.07
     บ
    0.06
     mund
    0.06
     베
    0.06
     Oilers
    0.06
     dö
    0.06
    돌
    0.06
    ipline
    0.06
    Act Density 0.006%

    No Known Activations