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

    Neuronpedia

    INDEX
    Explanations

    references to writing or typing in various contexts

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

    saying or writing

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    principalTable
    -0.47
    TabStop
    -0.45
    archiviato
    -0.43
    modelBuilder
    -0.42
     Вікі
    -0.41
    StructEnd
    -0.40
    どうやら
    -0.40
    featureID
    -0.39
     hierv
    -0.39
     daarvoor
    -0.37
    POSITIVE LOGITS
     saying
    0.52
     say
    0.48
     sagen
    0.47
     wrote
    0.46
     decir
    0.45
     written
    0.45
     écrit
    0.45
     Saying
    0.45
     dizer
    0.45
     écri
    0.45
    Act Density 0.188%

    No Known Activations