{"id":"75280117-4b00-4e6c-b4c6-fdb67ee284af","arxiv_id":"2501.04115","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For every prime p other than 3, this paper builds two large families of five-term permutation polynomials over F_{q^2}, unifying 76 earlier special cases and solving an open problem.","lead":"This paper constructs two large families of five-term permutation polynomials over finite fields F_{q^2}, valid in every characteristic other than 3. It unifies 76 previously scattered results and resolves an open problem for odd characteristics.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.1's proof misprints the gcd condition in the q≡1 branch: it uses gcd(Q−R+S,q+1) where the argument requires gcd(Q+R+S,q+1), so the proof as written fails for that branch, though the theorem statement is correct.","rationale":"The reader's conditional verdict is appropriate. The most load-bearing defect is the proof-text error in Section 3: the q≡1 mod3 branch of Theorem 1.1's proof replaces the correct exponent Q+R+S by Q−R+S when translating 'X^{Q+R+S} permutes μ_{q+1}' into a gcd condition. This is not a mere typographical curiosity: the wrong condition changes the answer for q=4, Q=1, R=4, S=8, and it is the exact step where Lemma 2.2's permutation criterion is converted into the theorem's arithmetic condition. The theorem statement itself has the correct Q+R+S, and the surrounding algebra (C1=U, C2=V, B_{3−z}/B_z equals g or 1/g) is consistent with the corrected condition. I verified the application of Lemma 2.2: with F_{q^2}, the subgroup is μ_{q+1}, and the lemma's exponent q−1 matches the second condition. Proposition 2.5 is proved from the cited lemmas and its use in both theorems is sound. The abstract's unqualified coefficient claim is also inaccurate when Q,R,S are not pairwise distinct, but this is a presentation issue. No deeper mathematical flaw surfaced beyond these presentation defects.","tokens_in":11987,"tokens_out":20403,"duration_ms":171897,"concrete_test":"Recompute the q≡1 branch of the proof of Theorem 1.1 using the exponent of X in g, which is Q+R+S, and replace the printed gcd(Q−R+S,q+1) with gcd(Q+R+S,q+1) in the sentence after 'or equivalently'. Then run a Sage check for q=4, Q=1, R=4, S=8, r=13: test whether f(X)=X^{13}(B_1(X^3)) with B_1=X^5−X^9−X^12+X^8−1 permutes F_{16}. The theorem's condition gives gcd(13,5)=1, the printed proof line gives gcd(5,5)=5, so the outcome distinguishes the correct condition.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3 reduces f to g(X)=ρ(X)∘X^{Q+R+S}∘η(X). For q≡1 mod3, ρ and η permute μ_{q+1}, so g permutes μ_{q+1} iff X^{Q+R+S} permutes μ_{q+1}, i.e. gcd(Q+R+S,q+1)=1. The manuscript instead says \"or equivalently gcd(Q−R+S,q+1)=1\". This is not harmless: for q=4, Q=1, R=4, S=8, gcd(Q−R+S,5)=gcd(5,5)=5, while gcd(Q+R+S,5)=gcd(13,5)=1. Since this line is the sole bridge from Lemma 2.2 to the permutation criterion in the q≡1 case, the proof as printed does not establish Theorem 1.1 for that case. The theorem statement itself has the correct condition, and the surrounding algebra (C1=U, C2=V, B_{3−z}/B_z equals g or 1/g) supports the corrected condition. I found no error in the application of Lemma 2.2 or Proposition 2.5; the latter is proved from the cited lemmas. A lesser presentation issue is the abstract's unqualified claim that coefficients lie in {1,−1}, which is false without the pairwise-distinct hypothesis; Theorem 1.1 itself includes that hypothesis.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs two large families of permutation pentanomials over F_{q^2}, of the form X^r B(X^{q-1}) where B(X) has at most five terms and (under a pairwise-distinctness hypothesis) coefficients in {1,-1}, for every prime p different from 3 and every q=p^k. Theorem 1.1 gives an if-and-only-if criterion in terms of gcd(r,q-1)=1 and gcd(Q+R+S,q+e)=1, where e is determined by q mod 3; Theorem 1.3 gives a second family requiring q≡1 mod 3 and gcd(Q-R+S,q+1)=1. The proofs use Zieve's reduction (Lemma 2.2) and a proposition about the Möbius maps ρ and η (Proposition 2.5). Theorem 1.7 provides F_q-linear equivalence reformulations when r=Q+R+S. The paper also includes tables identifying 76 prior results and conjectures from the characteristic-2 literature as instances of the new families, and notes implicit corrections to three of them.","tokens_in":12251,"tokens_out":4492,"duration_ms":41673,"significance":"If the results are correct, this is a substantial contribution: it unifies and vastly generalizes 76 existing characteristic-2 results and conjectures, and it provides the first permutation pentanomials of this shape in odd characteristic, resolving an open problem (Open Problem 2 of [5]). The proofs are refreshingly short and conceptual, avoiding the heavy case computations of the earlier literature, and the if-and-only-if form of the criteria is clean and falsifiable. The explicit subsumption tables and the stated corrections to prior results are valuable scholarly service. The only substantive flaw I found is a typo in the proof of Theorem 1.1 that is local and easily fixed; the theorem statements themselves have the correct gcd conditions.","major_comments":[{"comment":"The line reading \"X^{Q+R+S} permutes μ_{q+1}, or equivalently gcd(Q−R+S,q+1)=1\" is incorrect: the correct condition is gcd(Q+R+S,q+1)=1. As printed, the proof does not establish the criterion for the q≡1 branch; for example, with q=4, Q=1, R=4, S=8, one has gcd(Q−R+S,5)=5 whereas gcd(Q+R+S,5)=1. The theorem statement (Theorem 1.1) has the correct condition, and the immediately preceding equivalence with X^{Q+R+S} permuting μ_{q+1} supports the corrected expression, so this is almost certainly a typographical error. Nevertheless it is load-bearing in the proof-as-written and must be corrected.","section":"Section 3, proof of Theorem 1.1, q≡1 mod 3 branch"}],"minor_comments":[{"comment":"The abstract and introduction state that B(X) has coefficients in {1,-1} without mentioning the hypothesis that Q, R, S are pairwise distinct. The theorem statements include this hypothesis, so the abstract should be qualified to avoid overstating the coefficient condition.","section":"Abstract and Section 1"},{"comment":"The proof of Theorem 1.7 omits the details of several \"routine verifications\" and asks the reader to check them. Since this theorem is not needed for the main permutation results, this is acceptable, but a short derivation or an appendix would improve reproducibility.","section":"Section 5"},{"comment":"Reference [16] (Wu, Yuan, Ding, Ma, arXiv:2209.04762) does not appear to be cited in the body of the paper. Please either cite it where relevant or remove it from the bibliography.","section":"References"},{"comment":"The phrase \"One sees by inspection that U(X)=C2(X) and V(X)=C1(X)\" is terse; a brief indication of how the identity follows from the definitions of C1 and C2 would help the reader verify this key identification.","section":"Section 3"}],"recommendation":"minor_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should read this one. Ding and Zieve give two four-parameter families of permutation pentanomials over F_{q^2} for every prime p≠3, with clean iff conditions in terms of gcds. For p>2 these are reportedly the first such families, and they answer an open problem. In characteristic 2 the results subsume 76 earlier results and conjectures from 20 papers, with a careful table that also corrects three published theorems. The proofs are short and rest on Zieve's reduction lemma (Lemma 2.2) plus a Möbius-map argument; I spot-checked the gcd criteria and they match the theorem statements. Theorem 1.7 adds linear-equivalence interpretations that explain why the families work, though its proof omits some routine details.\n\nThe main soft spot is a typo in the proof of Theorem 1.1. In the q≡1 mod 3 branch it says \"gcd(Q−R+S, q+1)=1\" where the surrounding argument and the theorem statement require \"gcd(Q+R+S, q+1)=1\". This is not merely cosmetic: that line is the bridge from Lemma 2.2 to the permutation criterion in that branch. The theorem statement itself has the correct condition, and the q≡2 branch is fine, so the error is fixable, but it must be corrected before publication. The abstract also overstates the coefficient claim by dropping the pairwise-distinct condition on Q, R, S; Theorem 1.1 includes that condition, so the abstract should be aligned. Both are presentation issues, not flaws in the stated results.\n\nThe reliance on published lemmas is appropriate, and the citation pattern is thorough rather than self-promotional. The corrections to three earlier theorems are a useful service to the community. The main claims appear sound to me.\n\nThis paper deserves a serious referee. It is aimed at finite-field and permutation-polynomial specialists, who will get real value from it, and it also serves as a useful survey of a scattered literature. I would accept it after a quick revision fixing the typo and the abstract. I would cite it in my own work if I were active in this area.","headline":"Genuinely new permutation pentanomial families in all odd characteristics, with one fixable proof typo; the paper should be refereed and accepted after minor revision.","tokens_in":12821,"tokens_out":2653,"would_cite":true,"duration_ms":25232,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11T06"],"pacs":[],"model":"deepseek-v4-flash","headline":"For every prime p other than 3, this paper gives two four-parameter families of five-term polynomials that permute the field F_{q^2} exactly when two simple gcd conditions hold, unifying 76 prior results and resolving an open problem for…","keywords":["permutation polynomials","pentanomials","finite fields","roots of unity","Möbius transformations","gcd criteria","characteristic two","open problem"],"falsifier":"Exhaustively evaluate $f(X)=X\\cdot B_1(X^4)$ over $F_{25}$ for the data $p=5$, $q=5$, $Q=5$, $R=1$, $S=25$, $r=1$, where $B_1$ comes from Theorem 1.1 and Table 1; the theorem predicts a permutation, so finding any repeated value would refute it.","tokens_in":11744,"feed_emoji":"🔢","tokens_out":9612,"duration_ms":80386,"temperature":0.7,"pith_summary":"The paper proves that two large families of sparse polynomials over the field $F_{q^2}$ are permutation polynomials (bijections of the field to itself) under a simple arithmetical condition. For every prime $p$ other than 3 and every power $q=p^k$, the authors give five-term polynomials $X^r B(X^{q-1})$ with $B$ having coefficients in $\\{1,-1\\}$, and show they permute $F_{q^2}$ if and only if two gcd conditions hold. This unifies 76 earlier results and conjectures, all of which lived in characteristic 2, and for the first time produces such polynomials in odd characteristic, resolving an open problem. The proof is short and computation-free, replacing earlier case-by-case computer verification with a single conceptual reduction to fractional linear maps on roots of unity.","feed_headline":"One gcd test replaces 76 case-by-case permutation proofs","feed_subtitle":"For fields of any size, a five-term polynomial permutes F_{q^2} exactly when two coprime conditions hold.","key_machinery":"The central object is the pair of Möbius transformations $\\rho(X)=(X-\\omega)/(-\\omega X+1)$ and $\\eta(X)=(X+\\omega)/(\\omega X+1)$, where $\\omega$ is an element of order 3 in $F_{q^2}^*$. Proposition 2.5 shows these maps either permute the $(q+1)$-th roots of unity $\\mu_{q+1}$ (when $q\\equiv 1 \\pmod 3$) or interchange that set with the projective line $P^1(F_q)$ (when $q\\equiv 2 \\pmod 3$). The proof composes these maps with the monomial $X^n$ to build the rational function $g=\\rho\\circ X^n\\circ\\eta$, and shows the ratio $B_{3-z}(X)/B_z(X)$ on $\\mu_{q+1}$ equals $g$ or $1/g$. By Lemma 2.2, permutation of $F_{q^2}$ is equivalent to permutation of $\\mu_{q+1}$ by a derived function, so the whole problem reduces to asking when a monomial permutes $\\mu_{q+1}$ or $P^1(F_q)$, which is exactly the stated gcd condition.","core_discovery":"The central discovery is a necessary-and-sufficient criterion for permutation pentanomials of the form $X^r B_z(X^{q-1})$ over $F_{q^2}$. Theorem 1.1 states that when $Q,R,S$ are powers of $p$ and $B_1,B_2$ are the five-term polynomials from Table 1, the polynomial permutes $F_{q^2}$ iff $\\gcd(r,q-1)=1$ and $\\gcd(Q+R+S, q+e)=1$, where $e=1$ if $q\\equiv 1 \\pmod 3$ and $e=-1$ if $q\\equiv 2 \\pmod 3$. Theorem 1.3 gives a second family that permutes iff $q\\equiv 1 \\pmod 3$ and $\\gcd(Q-R+S, q+1)=1$. Because the arguments go through for every $p\\neq 3$, the special role of characteristic 2 in all 76 earlier results is shown to be an artifact of the literature, not of the mathematics.","pith_inferences":["The same fractional-linear reduction may apply to other sparse shapes: the proof only needs the ratio of the two 'companion' polynomials to be a conjugate of a monomial, so trinomials or heptanomials built from the same $\\omega$ may satisfy analogous gcd criteria.","Characteristic 3 is excluded because no order-3 element $\\omega$ with the required properties exists there; a modified construction using a different root-of-unity subgroup might fill that gap, though the present methods do not.","A companion result promised by the authors would likely complete a systematic classification of low-weight permutation polynomials of the form $X^r B(X^{q-1})$, with the two papers together covering the full landscape of such constructions."],"forward_implications":["Only the gcd conditions need checking: for any allowed $q$, $Q$, $R$, $S$ and any $r$ congruent to $Q+R+S$ mod $q+1$, the same two-line test decides whether the corresponding pentanomial permutes $F_{q^2}$.","All 76 prior permutation trinomial and pentanomial results and conjectures in characteristic 2 listed in Tables 3–6 are special cases of the two theorems, and three of them receive implicit corrections to their stated conditions.","Odd-characteristic permutation pentanomials of this form now exist for every prime other than 3, resolving the open problem noted in the literature.","When $r=Q+R+S$, the permutation is $F_q$-linearly equivalent to a simple monomial map on $F_{q^2}$ or on $F_q\\times F_q$, giving a one-page alternative proof for that case and subsuming the two longest prior papers on the topic."],"supporting_citations":[{"why":"Supplies Lemma 2.2, the reduction that turns permutation on $F_{q^2}$ into permutation on the $(q+1)$-th roots of unity; both main proofs start from it.","marker":"[21]"},{"why":"Supplies Lemmas 2.3 and 2.4, the fractional-linear bijection facts used in Proposition 2.5 to determine how the transformations $\\rho$ and $\\eta$ behave modulo 3.","marker":"[22]"},{"why":"A very recent characteristic-2 result that the paper's Theorem 1.7 subsumes in one-third of a page; the tables use it as a comparison point and correct a condition in it.","marker":"[5]"},{"why":"A recent 31-page characteristic-2 paper whose results Theorem 1.7 subsumes; it anchors the claim that the new proof is short.","marker":"[19]"},{"why":"A characteristic-2 paper whose results are subsumed and whose Table 2(14) receives an implicit correction (k odd); representative of the 76 subsumed results.","marker":"[20]"}],"fun_headline_variants":["One gcd rule unifies 76 permutation cases","Five-term permuters pinned by a simple gcd","Two gcds decide all permutation pentanomials","From 76 proofs to one gcd condition","For every prime but 3, a single gcd test"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything rests on the lemma that a polynomial $X^r B(X^{q-1})$ permutes $F_{q^2}$ exactly when $\\gcd(r,q-1)=1$ and the associated function permutes the $(q+1)$-th roots of unity; if that reduction fails in any case, the 'if and only if' conclusions of both theorems fail with it.","fun_headline_variants_meta":{"raw":{"variants":["One gcd rule unifies 76 permutation cases","Five-term permuters pinned by a simple gcd","Two gcds decide all permutation pentanomials","From 76 proofs to one gcd condition","For every prime but 3, a single gcd test"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000755,"raw_usage":{"total_tokens":3328,"prompt_tokens":885,"completion_tokens":2443,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":501,"completion_tokens_details":{"reasoning_tokens":2370}},"tokens_in":501,"tokens_out":2443,"duration_ms":17741,"temperature":1.0,"reasoning_tokens":2370,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T21:41:41.800076+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Exhaustively evaluate $f(X)=X\\cdot B_1(X^4)$ over $F_{25}$ for the data $p=5$, $q=5$, $Q=5$, $R=1$, $S=25$, $r=1$, where $B_1$ comes from Theorem 1.1 and Table 1; the theorem predicts a permutation, so finding any repeated value would refute it.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies Lemma 2.2, the reduction that turns permutation on $F_{q^2}$ into permutation on the $(q+1)$-th roots of unity; both main proofs start from it."},{"cited_title":"Some permutation pentanomials over finite fields of even characteristic","cited_arxiv_id":"2412.14641","evidence_quote":"A very recent characteristic-2 result that the paper's Theorem 1.7 subsumes in one-third of a page; the tables use it as a comparison point and correct a condition in it."},{"cited_title":"Zhang, H","cited_arxiv_id":null,"evidence_quote":"A recent 31-page characteristic-2 paper whose results Theorem 1.7 subsumes; it anchors the claim that the new proof is short."},{"cited_title":"Zhang, L","cited_arxiv_id":null,"evidence_quote":"A characteristic-2 paper whose results are subsumed and whose Table 2(14) receives an implicit correction (k odd); representative of the 76 subsumed results."}],"review_version":1}