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

    Neuronpedia

    INDEX
    Explanations

    reported or alleged actions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     sabe
    1.04
     knows
    0.97
     understands
    0.93
     compréhension
    0.91
     understand
    0.89
     compreender
    0.89
     entiende
    0.88
     जानते
    0.86
    理解
    0.86
     compreensão
    0.85
    POSITIVE LOGITS
     reportedly
    2.33
     allegedly
    1.85
     purportedly
    1.68
     announced
    1.67
     criticized
    1.65
     कथित
    1.64
     declared
    1.58
     protested
    1.58
     reported
    1.54
     repeatedly
    1.48
    Act Density 0.090%

    No Known Activations