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

    Neuronpedia

    INDEX
    Explanations

    referential phrases, particularly the word "which."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    introduces descriptive clauses

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     صوتيه
    -0.55
     चीज़ों
    -0.52
    -0.50
    adaptiveStyles
    -0.48
    InstrumentedTest
    -0.48
     okuyayım
    -0.46
    setVerticalGroup
    -0.46
     PagesJaunes
    -0.45
    цездатний
    -0.44
     coloured
    -0.42
    POSITIVE LOGITS
     שוליים
    0.56
    PMailer
    0.50
     often
    0.50
     חיצוניים
    0.48
    Which
    0.47
    errHandler
    0.46
    which
    0.45
     Which
    0.45
    often
    0.44
    middels
    0.44
    Act Density 0.072%

    No Known Activations