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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    This neuron detects tokens that serve as operands of C++ stream operators (the items immediately following `>>` or `<<` in `std::cin`/`std::cout` expressions).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    iyi
    -0.08
     içer
    -0.07
     scape
    -0.06
     한번
    -0.06
    К
    -0.06
     گفته
    -0.06
     pred
    -0.06
    VARCHAR
    -0.06
    (j
    -0.06
    ために
    -0.06
    POSITIVE LOGITS
    upos
    0.07
     حدود
    0.07
     zk
    0.07
    	tests
    0.07
    )?↵
    0.07
     Registrar
    0.06
    .chunk
    0.06
    (entries
    0.06
     booths
    0.06
     turtle
    0.06
    Act Density 0.118%

    No Known Activations