If you just want a raw count don’t use values and change aggfunc to ‘size’

df.pivot_talbe(index=['Pclass', 'Survived'], columns='Sex', aggfunc='size')