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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    code-like identifiers and technical jargon strings, especially mixed-case/camelCase or concatenated tokens.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
    0
    -0.10
    1
    -0.10
    5
    -0.10
    3
    -0.10
    4
    -0.09
    7
    -0.09
    9
    -0.09
    2
    -0.09
    13
    -0.08
    19
    -0.08
    POSITIVE LOGITS
     Birch
    0.08
    โล
    0.07
    odu
    0.07
    usta
    0.07
     Schwarz
    0.07
    enberg
    0.07
    uong
    0.07
    nore
    0.07
     Royale
    0.07
    staw
    0.07
    Act Density 4.068%

    No Known Activations