Part III · Quiz needs review

Quiz: Part III

Five questions on the chapters of Part III written so far. More will be added with the remaining chapters.

These questions ask you to combine ideas from several chapters. Choose an answer to see whether it is right and why. Your best score is kept in this browser only.

  1. Why does the formalization work with a smoothed prime sum?

    • Smoothing removes the zeros
    • After smoothing the amplitudes are about 2/γ22/\gamma^{2}, so the sum over all zeros converges absolutely and can be bounded
    • Lean cannot handle step functions

    The total of all amplitudes is about 0.0462 in units of x\sqrt{x}. Without smoothing the corresponding sum diverges.

  2. Under the hypothesis the scaled error stays within ±0.0462\pm 0.0462. What is that number?

    • A measured maximum from the lab
    • The first zero's amplitude
    • The sum of the amplitudes of all the waves: the worst case, in which every wave peaks at once

    It is ∑2/(14+γ2)\sum 2/(\tfrac14 + \gamma^{2}) over all zeros, equal to 2+γEuler−ln⁡4π2 + \gamma_{\text{Euler}} - \ln 4\pi. The first zero alone contributes 0.0100.

  3. The formalization proves only an upper bound. Why does that exclude zeros off the line?

    • An off-line zero adds an oscillation of size xβx^{\beta} that takes both signs, and by Landau's theorem the other terms cannot cancel it, so it would break an upper bound too
    • Because the error is always negative
    • It does not; a lower bound is still needed

    Lab 20.2 shows it: with β>12\beta \gt \tfrac12 the curve leaves the band through the top and the bottom.

  4. A Lean development compiles, but one lemma that the main theorem uses is closed with sorry. What has been proved?

    • The main theorem
    • The main theorem on the assumption that the lemma is true
    • Nothing at all

    Lean accepts sorry in place of any proof and records the dependency. Everything else is checked; the result is conditional until the gap is filled. This is the situation of trace_identity in this site's own Lean experiments.

  5. How does this proof's strategy differ in kind from Masullo's?

    • Both use a trace formula
    • This one counts zeros by computer
    • This one bounds a prime sum from one side and converts the bound into a statement about zeros; Masullo's compares the two sides of a trace formula

    They share only the fundamentals of Part I: the explicit link between a zero's real part and the size of its wave.