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

    Neuronpedia

    INDEX
    Explanations

    before start writing

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     examin
    0.45
     very
    0.44
     perceive
    0.44
     examine
    0.43
     disag
    0.42
     amplit
    0.41
     nonlinear
    0.41
     projective
    0.40
     decayed
    0.40
     unusual
    0.40
    POSITIVE LOGITS
     snag
    0.75
     graced
    0.64
     slapped
    0.61
    偷偷
    0.61
     donning
    0.60
     wr
    0.59
     sneak
    0.59
     nab
    0.59
     dutiful
    0.59
     whip
    0.58
    Act Density 0.046%

    No Known Activations