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

    Neuronpedia

    INDEX
    Explanations

    huge scale impressive quantities

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     полностью
    1.17
     totalmente
    1.12
     повністю
    1.03
    完全
    1.02
     completely
    1.02
     completamente
    1.01
     totally
    0.98
     entièrement
    0.95
     complètement
    0.95
     totalement
    0.94
    POSITIVE LOGITS
     astounding
    2.63
     incredible
    2.59
     remarkable
    2.53
     astonishing
    2.50
     extraordinary
    2.28
     tremendous
    2.25
     immense
    2.19
    incredible
    2.19
     unbelievable
    2.14
     staggering
    2.11
    Act Density 0.487%

    No Known Activations