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

    Neuronpedia

    INDEX
    Explanations

    Quotation marks

    np_max-act · gemini-2.0-flash

    This neuron detects standout thematic or emphatic keywords—often uncommon abstract nouns or titles—especially when they’re set off in quotes or headings.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Med
    -0.07
    -tracking
    -0.07
    -year
    -0.07
     er
    -0.07
    _it
    -0.07
     mientras
    -0.07
     client
    -0.07
    er
    -0.07
     departed
    -0.07
     end
    -0.07
    POSITIVE LOGITS
     Hà
    0.08
    :
    0.08
    urus
    0.07
     "',
    0.07
    ?.
    0.07
    aise
    0.07
    aux
    0.07
    _:
    0.07
    .
    0.07
     křes
    0.07
    Act Density 0.103%

    No Known Activations