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

    Neuronpedia

    INDEX
    Explanations

    than usual or expected

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    0.39
       
    0.32
     earum
    0.31
     এমনই
    0.31
    mselves
    0.30
    Approximately
    0.28
    Requirement
    0.28
     deoarece
    0.28
     همچنین
    0.28
    0.28
    POSITIVE LOGITS
     than
    0.51
     expected
    0.51
     usual
    0.51
     nötig
    0.47
    expected
    0.47
     anticipated
    0.44
     pure
    0.43
     they
    0.43
     kuin
    0.43
     gedacht
    0.43
    Act Density 0.080%

    No Known Activations