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

    Neuronpedia

    INDEX
    Explanations

    Shakespearean plays/poems

    np_max-act · gemini-2.0-flash

    The neuron responds most strongly to sentence- or clause-initial capitalized function words (e.g. “Of,” “For,” “And,” “First,” “Upon”)—i.e. discourse markers at the start of sentences.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    /aws
    -0.07
    verts
    -0.06
     vis
    -0.06
     Nine
    -0.06
    nodes
    -0.06
     manga
    -0.06
     Mods
    -0.06
    ị
    -0.06
     cpu
    -0.06
     необходимости
    -0.06
    POSITIVE LOGITS
     wi
    0.07
     QR
    0.07
    achelor
    0.07
    Boy
    0.06
     wei
    0.06
    ANO
    0.06
    _sq
    0.06
    bohydr
    0.06
     Engine
    0.06
     base
    0.06
    Act Density 0.020%

    No Known Activations