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

    Neuronpedia

    INDEX
    Explanations

    Numerical data types

    np_max-act · gemini-2.0-flash

    JSON-related syntax and types used in programming interfaces.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    This neuron activates on tokens that delimit code blocks or inline code (e.g. backticks and other code‐boundary markers).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Matlab
    -0.07
    iliz
    -0.06
     Lev
    -0.06
     architekt
    -0.06
     Muk
    -0.06
     fatigue
    -0.06
     Being
    -0.06
    CONS
    -0.06
     disk
    -0.06
     rap
    -0.06
    POSITIVE LOGITS
     eds
    0.07
     VERY
    0.07
     fraught
    0.07
    ，在
    0.06
    애
    0.06
     tuto
    0.06
     zprac
    0.06
     contention
    0.06
     inadvertently
    0.06
     discrepancy
    0.06
    Act Density 0.013%

    No Known Activations