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

    Neuronpedia

    INDEX
    Explanations

    code snippets

    np_max-act · gemini-2.0-flash

    The neuron strongly activates on long runs of the same token repeated over and over, i.e. monotonous repetition sequences.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Illum
    -0.07
    �
    -0.07
     حذ
    -0.07
    �
    -0.07
     Exterior
    -0.07
    亜
    -0.06
     �
    -0.06
     �
    -0.06
    릴
    -0.06
    &display
    -0.06
    POSITIVE LOGITS
     CITY
    0.07
     hearty
    0.07
    932
    0.06
    ój
    0.06
    dirname
    0.06
    AMENT
    0.06
    内の
    0.06
     그래서
    0.06
     Admission
    0.06
    @Repository
    0.06
    Act Density 0.017%

    No Known Activations