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

    Neuronpedia

    INDEX
    Explanations

    version

    np_max-act · gemini-2.0-flash

    This neuron detects numeric version literals (floating-point numbers), as used for license or version numbers in headers.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .serialize
    -0.06
    _learning
    -0.06
     BX
    -0.06
     Firewall
    -0.06
    _players
    -0.06
     Мас
    -0.06
    리
    -0.06
     Universe
    -0.06
    \grid
    -0.06
    _deleted
    -0.06
    POSITIVE LOGITS
     interfaces
    0.07
    іп
    0.07
     aby
    0.07
    Picture
    0.06
    ategories
    0.06
    	rc
    0.06
     iam
    0.06
     dice
    0.06
     ord
    0.06
    robot
    0.06
    Act Density 0.003%

    No Known Activations