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

    Neuronpedia

    INDEX
    Explanations

    vector length/normalization

    np_max-act · gemini-2.0-flash

    The neuron is detecting numeric literals (in particular floating‐point numbers) in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ektiv
    -0.06
     дод
    -0.06
     pap
    -0.06
     двор
    -0.06
    rnd
    -0.06
     Buffy
    -0.06
     Khu
    -0.06
     folder
    -0.06
    _CMP
    -0.06
     projektu
    -0.06
    POSITIVE LOGITS
    ';↵↵↵
    0.07
    Specify
    0.07
    NODE
    0.07
     integr
    0.06
    ----------↵↵
    0.06
     Rotate
    0.06
     subscriptions
    0.06
     Jensen
    0.06
    .amount
    0.06
    /example
    0.06
    Act Density 0.005%

    No Known Activations