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

    Neuronpedia

    INDEX
    Explanations

    conversational prompts

    np_max-act · gemini-2.0-flash

    This neuron fires on instructional or procedural language—particularly step-by-step directions phrased with action verbs (e.g. “draw,” “ask,” “have,” “write,” etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    assuming
    -0.07
    ٬
    -0.07
    uned
    -0.07
    攝
    -0.06
    _PROPERTY
    -0.06
     تیم
    -0.06
     commissioned
    -0.06
    	rec
    -0.06
     thinner
    -0.06
     OBS
    -0.06
    POSITIVE LOGITS
     Calif
    0.07
     willen
    0.06
     yıldır
    0.06
    elu
    0.06
    izada
    0.06
     Доб
    0.06
     aucun
    0.06
     zeigen
    0.06
     relation
    0.06
     จำก
    0.06
    Act Density 0.131%

    No Known Activations