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

    Neuronpedia

    INDEX
    Explanations

    positive evaluation of code

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     কিছুটা
    0.49
     complicates
    0.48
     somewhat
    0.47
    必要があります
    0.47
     tricky
    0.45
    decrypt
    0.45
    大致
    0.45
     craziness
    0.45
     vagu
    0.44
     necesitamos
    0.43
    POSITIVE LOGITS
     Tested
    0.65
    Excellent
    0.60
     Excellent
    0.59
     thoroughly
    0.57
     excellent
    0.56
     tested
    0.55
     Excelente
    0.55
    Tested
    0.54
     excelentes
    0.52
     ausgezeichnet
    0.52
    Act Density 0.001%

    No Known Activations