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

    Neuronpedia

    INDEX
    Explanations

    Instructions/procedures

    np_max-act · gemini-2.0-flash

    The neuron detects patent‐style preference or advantage statements—words and phrases (e.g. “preferably,” “it will be of advantage,” “are employed to”) that introduce preferred embodiments or recommended features.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     багатьох
    -0.07
    Phrase
    -0.07
     rz
    -0.07
    ��
    -0.07
     Integrity
    -0.06
     유저
    -0.06
    tolower
    -0.06
     vůbec
    -0.06
     právo
    -0.06
     Enables
    -0.06
    POSITIVE LOGITS
    cerer
    0.08
    auté
    0.07
     QLabel
    0.07
     установ
    0.06
    obile
    0.06
    arten
    0.06
    яет
    0.06
     textField
    0.06
    gary
    0.06
     Moines
    0.06
    Act Density 0.115%

    No Known Activations