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

    Neuronpedia

    INDEX
    Explanations

    ion

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on fragments of personal names—especially uncommon or foreign‐sounding surnames.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    """
↵
    -0.07
     verw
    -0.06
     mus
    -0.06
    (part
    -0.06
    *z
    -0.06
     obra
    -0.05
     उम
    -0.05
    hart
    -0.05
    Sus
    -0.05
     δεδο
    -0.05
    POSITIVE LOGITS
    ropping
    0.08
     Οικο
    0.07
     Κο
    0.07
    öh
    0.07
    connected
    0.07
     왜
    0.06
    .met
    0.06
    Driver
    0.06
    отв
    0.06
    trinsic
    0.06
    Act Density 0.076%

    No Known Activations