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

    Neuronpedia

    INDEX
    Explanations

    words related to the concept of "speaking" or "speech."

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

    SPELL\_

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Touchable
    -0.53
     Cotter
    -0.52
     Dund
    -0.52
    uint
    -0.50
    matrix
    -0.50
    Życiorys
    -0.49
     Caron
    -0.48
     Incon
    -0.47
    Ligações
    -0.46
     Margin
    -0.46
    POSITIVE LOGITS
     Spe
    2.02
    Spe
    1.98
     spe
    1.77
    spe
    1.76
     SPE
    1.54
    SPE
    1.45
     spé
    1.06
     Speer
    0.86
     espe
    0.85
     Spear
    0.85
    Act Density 0.004%

    No Known Activations