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

    Neuronpedia

    INDEX
    Explanations

    aggregate

    np_max-act · gemini-2.0-flash

    This neuron detects occurrences of the root “aggregat” (in words like aggregate, aggregated, aggregation).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ンブ
    -0.07
     courtesy
    -0.07
     foresee
    -0.07
     fits
    -0.07
     Teknik
    -0.07
     costumes
    -0.06
     Trou
    -0.06
     fit
    -0.06
     prisoners
    -0.06
     Sü
    -0.06
    POSITIVE LOGITS
     aggregate
    0.12
     Aggregate
    0.12
    aggregate
    0.11
     aggregates
    0.11
     aggregation
    0.11
    Aggregate
    0.09
     aggregated
    0.09
     aggreg
    0.08
    _ag
    0.08
     aggregator
    0.08
    Act Density 0.014%

    No Known Activations