Some Remarks on Oscillation Inequalities

The paper proves a uniform oscillation inequality for ergodic polynomial averages with constants depending only on d, k, p and deg P and independent of coefficients, via a detailed major/minor-arcs and transference argument (Theorem 1.4, with uniformity explicitly stated) . By contrast, the model’s proof hinges on a claimed “oscillation ≤ jump” reduction: O^2 ≤ C·J_2. This step is not established in the paper and is, in fact, contradicted by the paper’s examples and theory. The authors show that oscillations and jumps are not comparable in the sense required—there exist functions for which r=2 jump bounds hold while r-oscillation bounds blow up (Remark 2.25), and they prove that oscillations cannot serve as endpoint surrogates for r-variations nor dominate jump counts in the required uniform way (Theorem 1.9 and Lemma 4.7) . Hence the model’s Step 0 is invalid, so its reduction from jumps to oscillations does not go through, even though the final theorem asserted by the model matches the paper’s correct result.