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

    Neuronpedia

    INDEX
    Explanations

    strong opinions

    np_max-act · gemini-2.0-flash

    This neuron detects negative descriptive adjectives—pejorative qualifiers like “dirty,” “strange,” or other insulting descriptors.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    words and short phrases that express strong subjective evaluation or emotional emphasis (especially negative judgments).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
     forsk
    -0.07
    Former
    -0.06
    澤
    -0.06
     khi
    -0.06
    _fsm
    -0.06
    _mas
    -0.06
    before
    -0.06
    _fee
    -0.06
     navCtrl
    -0.06
     lamin
    -0.06
    POSITIVE LOGITS
    ubb
    0.07
     Kiev
    0.07
     offspring
    0.07
     trứng
    0.06
     Treatment
    0.06
     Q
    0.06
     Counter
    0.06
     Views
    0.06
    _Show
    0.06
     +++
    0.06
    Act Density 0.118%

    No Known Activations