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

    Neuronpedia

    INDEX
    Explanations

    Medical abnormalities

    np_max-act · gemini-2.0-flash

    This neuron detects the subordinating adverb “where” (especially in the phrase “where the … is …”) marking the start of a relative clause.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ipelines
    -0.07
    -ton
    -0.07
    еф
    -0.07
    乐
    -0.07
    Disney
    -0.07
     мус
    -0.06
    Jet
    -0.06
     billionaires
    -0.06
    dart
    -0.06
    ело
    -0.06
    POSITIVE LOGITS
     Latest
    0.07
     Pond
    0.06
     assists
    0.06
     Height
    0.06
     Integrated
    0.06
     hamstring
    0.06
     schwar
    0.06
    pri
    0.06
     Join
    0.06
     jails
    0.06
    Act Density 0.089%

    No Known Activations