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

    Neuronpedia

    INDEX
    Explanations

    specifications

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on numeric measurements with decimal points (e.g. specification values), i.e. floating-point figures in technical descriptions.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     ol
    -0.07
    Ra
    -0.07
    .Toast
    -0.07
    _LD
    -0.07
    _SENT
    -0.06
    WebHost
    -0.06
     Ra
    -0.06
    rb
    -0.06
     Hand
    -0.06
    Cas
    -0.06
    POSITIVE LOGITS
     specifications
    0.14
     specification
    0.13
     Specification
    0.09
     spec
    0.09
    Specification
    0.08
     Specifications
    0.08
     scripture
    0.08
    isión
    0.08
     specs
    0.07
    .localScale
    0.07
    Act Density 0.005%

    No Known Activations