Boundedness of the sequence n^n^(-x)

  • Thread starter bins4wins
  • Start date
  • Tags Tags
    Sequence
3 replies · 2K views · last reply
bins4wins
Messages
14
Reaction score
0
For any real ##x > 0##, prove that the sequence ##n^{n^{-x}}## is bounded (and if possible, monotonically decreasing after some point). The catch is that logarithms and the exponential constant cannot be used. We must arrive at the proof using fairly "primitive tools"
If you look at the graph of the function ##f(n) = n^{n^{-x}}## you'll notice that it increases at first but then decreases asymptotically towards 1 after some point. My attempt at a solution consisted of choosing three integers ##a < b < c##, then assuming that if ##a^{a^{-x}} > b^{b^{-x}}## then ##b^{b^{-x}} > c^{c^{-x}}##, but I'm not sure if there is enough information in the hypothesis to prove the desired.
 
Last edited:
bins4wins said:
For any real ##x > 0##, prove that the sequence ##\large n^{n^{-x}}## is bounded (and if possible, monotonically decreasing after some point). The catch is that logarithms and the exponential constant cannot be used. We must arrive at the proof using fairly "primitive tools"

If you look at the graph of the function $f(n) = n^{n^{-x}}$ you'll notice that it increases at first but then decreases asymptotically towards 1 after some point. My attempt at a solution consisted of choosing three integers $a < b < c$, then assuming that if $a^{a^{-x}} > b^{b^{-x}}$ then $b^{b^{-x}} > c^{c^{-x}}$, but I'm not sure if there is enough information in the hypothesis to prove the desired.
Hello bins4wins. Welcome to PF !

Use double $, $​$, for LaTeX tags

or

use double #, #​#, for inline LaTeX tags

(I used double # in the first line of the above quoted message.)