The large-characteristic Newton polygon conjecture for twisted binomial L-functions

About 5 years old · traced to

Let χ\chi be a multiplicative character of order cc over a finite field, let f(x)=xd+λxef(x)=x^d+\lambda x^e with (d,e)=1(d,e)=1, and let pp be the characteristic. Write NP⁡u,m(f)\operatorname{NP}_{u,m}(f) and NP⁡u,T(f)\operatorname{NP}_{u,T}(f) for the twisted classical and TT-adic Newton polygons, respectively, and let Pu,e,dP_{u,e,d} be the associated lower-bound polygon. Large-characteristic Newton polygon conjecture. If pp is large enough with respect to cc and dd, then

NP⁡u,m(f)=NP⁡u,T(f)=Pu,e,d.\operatorname{NP}_{u,m}(f)=\operatorname{NP}_{u,T}(f)=P_{u,e,d}.

This conjecture generalizes the conjecture of Zhang and Niu and is motivated by the criterion that equality holds when the associated integral constant is not divisible by pp. The paper proves the equality for p>c(d2−d+1)p>c(d^2-d+1) when e=d−1e=d-1, but the general large-characteristic assertion remains open.

References

Primary source

Shenxing Zhang, “On the Newton polygons of twisted L-functions of binomials”, arXiv:2109.14852 (2021).

Progress summary

Refreshed
Open

The conjecture remains open: only a restricted family of cases has been proved, with no verified resolution of the general claim.

The conjecture asserts that, for sufficiently large characteristic, the twisted classical and TT-adic Newton polygons both equal the predicted lower-bound polygon. Shenxing Zhang stated it as Conjecture 1.5 in a preprint posted on September 30, 2021, for binomials f(x)=xd+λxef(x)=x^d+\lambda x^e.

Known results

  • Zhang, 2021: if p>(d−e)(2d−1)p>(d-e)(2d-1), both Newton polygons lie above Pu,e,dP_{u,e,d}.
  • Zhang, 2021: equality holds exactly when the associated integral constant Hμ,c,p,e,dH_{\mu,c,\mathbf p,e,d} is not divisible by pp.
  • Zhang, 2021: if e=d−1e=d-1 and p>c(d2−d+1)p>c(d^2-d+1), then NP⁡u,m(f)=NP⁡u,T(f)=Pu,e,d\operatorname{NP}_{u,m}(f)=\operatorname{NP}_{u,T}(f)=P_{u,e,d}.

September 30, 2021 preprint

Zhang’s preprint establishes the lower bound and the case e=d−1e=d-1, but explicitly leaves the general sufficiently-large-pp assertion as a conjecture. The retrieved record contains no later proof, counterexample, or claimed settlement.

Current status (as of September 2026): The conjecture is unproved in general; the equality is settled for e=d−1e=d-1 under p>c(d2−d+1)p>c(d^2-d+1) and related special cases, while the general large-characteristic assertion remains open.

Sources

Solutions 0

No solutions have been posted yet.