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

    Neuronpedia

    INDEX
    Explanations

    Code and passwords

    np_max-act · gemini-2.0-flash

    The neuron activates on long, dense alphanumeric strings (often with symbols or padding) that resemble secrets such as passwords, keys, or tokens.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     talk
    -0.07
    -To
    -0.07
    상위
    -0.06
    яз
    -0.06
     fds
    -0.06
    ogenerated
    -0.06
    Ts
    -0.06
    _home
    -0.06
     Noah
    -0.06
    ESS
    -0.06
    POSITIVE LOGITS
     Поэтому
    0.06
    Ђ
    0.06
     Strait
    0.06
    �
    0.06
    ンテ
    0.06
     phố
    0.06
    atur
    0.06
    IVAL
    0.06
     Rotary
    0.06
    คราม
    0.06
    Act Density 0.011%

    No Known Activations