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

    Neuronpedia

    INDEX
    Explanations

    attention mechanism

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Germanic
    0.50
     reeds
    0.49
     chromatic
    0.47
     leiden
    0.46
    ersion
    0.45
     complexion
    0.44
     rustic
    0.44
     astring
    0.44
     rambling
    0.44
     pleated
    0.42
    POSITIVE LOGITS
     ఇటీవల
    0.55
    อร์
    0.48
     компанию
    0.48
     предотвра
    0.47
    বুব
    0.47
    Coronavirus
    0.46
     హీరో
    0.45
    خدم
    0.45
    шим
    0.45
    Voice
    0.45
    Act Density 0.003%

    No Known Activations