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

    Neuronpedia

    INDEX
    Explanations

    =

    np_max-act · gemini-2.0-flash

    This neuron detects column‐to‐column equality comparisons (the “=“ in SQL JOIN … ON conditions).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     circle
    -0.07
    AYS
    -0.06
    _PRESS
    -0.06
     光
    -0.06
     CRS
    -0.06
    ิน
    -0.06
     emulator
    -0.06
     Iris
    -0.06
     pathway
    -0.06
     두
    -0.06
    POSITIVE LOGITS
    ('-',
    0.06
    цев
    0.06
     Σε
    0.06
     склада
    0.06
    ляет
    0.06
    ätzlich
    0.06
    获得
    0.06
    >Status
    0.06
     valeurs
    0.06
     Seamless
    0.06
    Act Density 0.003%

    No Known Activations