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

    Neuronpedia

    INDEX
    Explanations

    anned

    np_max-act · gemini-2.0-flash

    This neuron detects the word “scanned” (as part of scanned‐page markers in image/file metadata).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    出去
    -0.08
     χωρίς
    -0.07
    번호
    -0.06
    .ds
    -0.06
    ічних
    -0.06
     '".$
    -0.06
     thấy
    -0.06
     svých
    -0.06
     OTP
    -0.06
    _rad
    -0.06
    POSITIVE LOGITS
     entrev
    0.07
    .Face
    0.07
    specs
    0.06
    528
    0.06
    Backing
    0.06
     thrust
    0.06
     discrim
    0.06
    .Quit
    0.06
     prac
    0.06
    crc
    0.06
    Act Density 0.004%

    No Known Activations