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

    Neuronpedia

    INDEX
    Explanations

    code/data

    np_max-act · gemini-2.0-flash

    The neuron flags tokens and punctuation inside the list of candidate relation labels (e.g., “Cause-Effect,” “Component-Whole,” etc.) and their associated score entries.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    _modes
    -0.07
     Diaz
    -0.07
    _ping
    -0.07
     transforming
    -0.06
     cadena
    -0.06
    Forum
    -0.06
    뢰
    -0.06
    CanBe
    -0.06
     IEEE
    -0.06
     κατα
    -0.06
    POSITIVE LOGITS
    perience
    0.06
     menor
    0.06
     이미지
    0.06
    umin
    0.06
    %↵↵
    0.06
    _ak
    0.06
     ENERGY
    0.06
     ног
    0.06
     chví
    0.06
    :last
    0.06
    Act Density 0.089%

    No Known Activations