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

    Neuronpedia

    INDEX
    Explanations

    preparation for input

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ಕ್ರ
    0.51
     зла
    0.47
     שנים
    0.47
     grievous
    0.46
    ڑ
    0.46
    ਣਾ
    0.46
    大きさ
    0.46
     ஆண்டுகள்
    0.45
     nerveux
    0.45
     خواہش
    0.45
    POSITIVE LOGITS
     pre
    0.52
     Before
    0.51
     before
    0.51
     prepare
    0.50
     initial
    0.49
     onboarding
    0.48
     input
    0.47
     preparing
    0.46
     prepared
    0.45
     during
    0.44
    Act Density 0.077%

    No Known Activations