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

    Neuronpedia

    INDEX
    Explanations

    't

    np_max-act · gemini-2.0-flash

    This neuron detects occurrences of the contraction “don’t” (i.e. the word “don’t”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    (A
    -0.08
    recognized
    -0.07
     (*((
    -0.07
    Ａ
    -0.06
     Judicial
    -0.06
     sea
    -0.06
     temple
    -0.06
     sums
    -0.06
     seaw
    -0.06
     reproduce
    -0.06
    POSITIVE LOGITS
     Dont
    0.09
     dont
    0.08
    dont
    0.08
     productName
    0.08
    ont
    0.08
    't
    0.07
    ’t
    0.07
     έχουν
    0.07
    นคร
    0.07
    俺
    0.07
    Act Density 0.023%

    No Known Activations