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

    Neuronpedia

    INDEX
    Explanations

    debugging/investigating errors

    np_max-act · gemini-2.0-flash

    This neuron specializes in spotting numeric tokens—especially version numbers, error codes, and other digits (including decimals) in technical logs.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Orchestra
    -0.07
     már
    -0.06
    cona
    -0.06
    [k
    -0.06
    Football
    -0.06
    ullen
    -0.06
    selector
    -0.06
    如此
    -0.06
    \DependencyInjection
    -0.06
     kennenlernen
    -0.06
    POSITIVE LOGITS
     Benefits
    0.06
     العربية
    0.06
     cil
    0.06
    _GPU
    0.06
     disagreements
    0.06
     cel
    0.06
     hl
    0.06
    vements
    0.05
     토
    0.05
     Executors
    0.05
    Act Density 0.048%

    No Known Activations