{"id":"9c53e9b3-1e02-488f-819f-9ead03dc34f2","arxiv_id":"2506.14966","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For p(z)=z^m+c(z^k+conj(z)^k)-1, the paper derives exact minimum and maximum zero counts as c varies, using zeros confined to 2m rays and real-valued polynomial counts per ray.","lead":"This paper proves exact formulas for how many zeros two families of complex-valued harmonic functions have as the parameter c changes. It matters because it extends zero-counting results for harmonic trinomials to a new family using a real-variable ray-counting method.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Most load-bearing concern: Sections 4.3–4.4 omit the derivations for all but one residue class in each theorem, leaving the central zero-count tables unverified as printed.","rationale":"The paper has a clear and mostly sound strategy: zeros lie on 2m rays, each ray reduces to a real function f_j, and Lemmas 2–11 correctly determine the number of positive zeros of f_j in each of the six relevant cases. The worked example in Section 4.3 (m≡1 mod 4, k even) is internally consistent, and the analogous argument in Section 4.4 (m≡3 mod 4, k odd) also checks out. My independent audit of the remaining residue classes, including the α=0 cases, reproduces every tabulated value of M and N in Theorems 1 and 2. The weakness is therefore not a detected mathematical error but an omission in the proof: the theorems are asserted for all cases, yet the derivations for most cases are left to the reader. In a field where exact zero-count theorems are the main result, this is a genuine completeness concern and justifies the reader's CONDITIONAL verdict. The reader's weakest_assumption identifies exactly this gap, and I agree with that assessment. No new objection beyond the reader's was found, so the verdict should remain unchanged.","tokens_in":10349,"tokens_out":18131,"duration_ms":169649,"concrete_test":"Run a systematic enumeration of the omitted cases: for each residue class of m mod 4 and each parity of k, use Propositions 1–3 to count the number of even and odd j in {0,...,2m-1} with Re(e^{ikj\\pi/m})>0, =0, and <0, substitute those counts into Lemmas 2–7 for k>0 or Lemmas 8–11 for k<0, and compare the resulting minimum and maximum zero counts with the tables in Theorems 1 and 2. A direct computational variant is to apply Sturm sequences to each f_j for representative pairs such as (m,k)=(12,5) and (13,6), for both positive and negative k, at c=0.01 and c=100; agreement with the theorems would confirm that the omitted cases are routine.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claims are exact zero-count transitions for p(z)=z^m+c(z^k+\\bar z^k)-1, and the proof reduces the problem to counting, for each j, the positive real zeros of f_j(r)=(-1)^j r^m+2c\\alpha r^k-1 with \\alpha=\\cos(kj\\pi/m). The lemmas in Section 3 are rigorous, but the theorems are only proved for one case each: Section 4.3 says 'A complete proof of the theorem would require many cases. We give the proof for one of the more complicated cases and leave the others to the reader,' and Section 4.4 says the same. The tabulated N and M values for the other residue classes depend on unstated counts of even and odd j for which Re(e^{ikj\\pi/m}) is positive, negative, or zero, combined with Lemmas 2–11. These counts are not routine in every case: for example, when m\\equiv 0 mod 4 and k is odd, \\alpha=0 occurs for some rays (e.g., m=12, k=5 gives j=6,18), and the fixed contribution of those rays must be handled separately. Because the theorems assert exact values for all m,k, the published proof is incomplete as written. I independently audited the omitted cases by counting roots of unity by parity and reproduced all stated M and N values, so this is a completeness gap rather than an identified error.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the zeros of the one-parameter families of complex-valued harmonic functions p(z)=z^m+c(z^k+\\bar z^k)-1, where c>0, gcd(m,k)=1, m>|k|, and k is either positive (Theorem 1) or negative (Theorem 2). The main structural result, Lemma 1, shows that every zero lies on one of the 2m rays with angle jπ/m, which reduces the problem to counting positive real zeros of the real functions f_j(r)=(-1)^j r^m+2cα r^k-1 with α=cos(kjπ/m). Section 3 proves six lemmas describing how many positive real zeros f_j has according to the sign of k, the parity of j, and the sign of α; the thresholds in the two-zeros cases are given explicitly. Section 4 then counts, for each residue class of m modulo 4 and each parity of k, how many even and odd j yield α positive, negative, or zero, and uses those counts with the lemmas to assert exact minimum and maximum zero counts as c varies. Theorems 1 and 2 state monotone transitions from m to m+N (for positive k) and from M+N to M (for negative k), with N and M tabulated by m mod 4 and the parity of k. The paper's proofs, however, work out only one case in each theorem and leave the remaining five cases in each theorem to the reader.","tokens_in":10601,"tokens_out":6754,"duration_ms":69661,"significance":"If the missing cases are supplied, the paper would provide a genuinely elementary and exact zero-counting result for two families whose critical curves are not circles, complementing the argument-principle methods of Brilleslyper et al. and Brooks-Lee. The ray-reduction idea is clean, the real-variable lemmas are correct, and the worked examples are consistent with the stated theorems. The thresholds c0 in Lemmas 7 and 11 are explicit and parameter-free. The main weakness is that the published proofs of Theorems 1 and 2 are incomplete: only one case in each theorem is actually proved, and the exact tabulated values for all other cases rest on unverified parity-and-sign counts of roots of unity. The omitted counts appear to be correct in the cases I checked, but a complete proof is needed before the theorems can be accepted as proven.","major_comments":[{"comment":"The proof explicitly states, 'A complete proof of the theorem would require many cases. We give the proof for one of the more complicated cases and leave the others to the reader.' The theorem has six cases depending on m mod 4 and the parity of k, but only the case m≡1 mod 4, k even is actually proved. The exact values of N in the theorem are computed from the asserted counts of even and odd j with Re(ω^j)>0, <0, or =0, and for the remaining five cases these counts are not established. These counts are load-bearing, not a routine repetition: for example, cases with m≡0 mod 4 and odd k have rays with α=0, whose contribution must be handled through Lemmas 2 and 3. As printed, the proof of Theorem 1 is therefore incomplete. The authors should either provide a complete case analysis or prove a general counting lemma that yields all six rows of the table.","section":"Section 4.3, proof of Theorem 1"},{"comment":"The same incompleteness occurs in the proof of Theorem 2. Only the case m≡3 mod 4, k odd is worked out, and the other five cases are left to the reader. The statement 'we give the proof for one of the more complicated cases and leave the others to the reader' does not supply the required counting argument for the remaining residue classes. In particular, for m≡0 mod 4 and odd k, there are rays with cos(kjπ/m)=0 (e.g., m=12, k=5 gives j=6,18), so the claimed 'similar strategy' must also account for the fixed contributions of those rays via Lemmas 2 and 3. Since the M and N values in Theorem 2 depend on these unproved counts, the theorem is not established as written. A complete case analysis or a uniform counting lemma with full proof is required.","section":"Section 4.4, proof of Theorem 2"},{"comment":"The monotonicity claim is only implicit. The authors show that for small c the count is m (or M+N) and for large c it is m+N (or M), and that on each ray with a threshold the count changes by two at a single c0. Since the per-ray counts are monotone in c, the total count is nondecreasing (Theorem 1) or nonincreasing (Theorem 2), but this aggregation should be stated explicitly, including the behavior at the threshold values where a double zero occurs. This is a presentation issue rather than a substantive error, but it should be clarified in the revision.","section":"Sections 4.3-4.4 and Lemmas 7, 11"}],"minor_comments":[{"comment":"In both proofs, the phrase 'j∈{0,2,...,2m−1}' should be 'j∈{0,1,...,2m−1}' to match Lemma 1 and the surrounding discussion.","section":"Section 4.3 and 4.4"},{"comment":"In the bullet for 'm≡1 mod 4 and k is even', there is a missing comma: it should read 'M=m, N=m+1'.","section":"Theorem 2"},{"comment":"There is a typo in the proof: 'postive real zeros' should be 'positive real zeros'.","section":"Lemma 3"},{"comment":"The displayed formulas should consistently use \\bar z^k for the conjugate term; in the plain-text rendering both terms appear as z^k, which obscures the conjugate.","section":"Examples 1 and 2"},{"comment":"Reference [9] to Sheil-Small is incomplete: it lists only a date and lacks a title, venue, or preprint identifier. Full bibliographic data should be supplied.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The core idea is sound and the omitted cases appear to be correct, so I do not suspect an error in the statements. The issue is that the proofs of the two main theorems are incomplete as printed, which is not acceptable for publication. The revision should add a complete case analysis or a general counting lemma for the roots-of-unity counts, and I would ask the handling editor to insist on this rather than allowing the 'other cases are left to the reader' language to stand."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe headline is that this is a genuine new result—exact zero-count transitions for the harmonic polynomials p(z)=z^m+c(z^k+\\bar{z}^k)-1, with positive and negative k—obtained through a ray-reduction trick that is quite different from the earlier hypocycloid/argument-principle methods. The two families do not appear in the cited literature, and the real-variable counting on the 2m rays is a clean idea that will interest the zero-location community.\n\nThe Section 3 lemmas are correct. I checked the sign cases, the critical point calculations, and the threshold constants; they all line up. The one fully worked example in each theorem is consistent. The citations are appropriate and I don't see self-citation padding. No parameters were fitted; the proof is self-contained from standard calculus and roots-of-unity facts.\n\nThe soft spot is exactly what the reader flagged: the proofs of Theorems 1 and 2 are not complete. Section 4.3 proves one residue class and then says the rest are 'left to the reader.' Section 4.4 does the same. For a theorem whose entire content is a table over residue classes, that's not acceptable. Some of the unstated cases are not routine—for m≡0 mod 4 and k odd, rays with α=0 appear and need separate counting. An independent audit of the omitted cases, in the stress-test note, reproduces every stated M and N, so this is a completeness gap rather than a discovered error. But as printed, the central theorems are under-proved.\n\nThis is for specialists. If you work on zero location for harmonic polynomials, you'll want to know this result and this method. I'd bring it to reading group maybe, and I'd cite it if I needed a new family with exact counts.\n\nRecommendation: send it to a serious referee. The referee should require the omitted case derivations, or at least a table giving the counts for every residue class, before acceptance. The paper's core seems right, but it isn't fully demonstrated yet.\n\nBest,","headline":"Solid new zero-counting result for two harmonic trinomial families, but the proof is incomplete as printed—most cases are deferred to the reader.","tokens_in":11151,"tokens_out":6317,"would_cite":true,"duration_ms":58826,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["30C15"],"pacs":[],"model":"deepseek-v4-flash","headline":"For $p(z)=z^m+c(z^k+\\bar z^k)-1$ with $\\gcd(m,k)=1$, the zero count changes monotonically with $c$, rising from $m$ to $m+N$ when $k>0$ and falling from $M+N$ to $M$ when $k<0$, with $M$ and $N$ read from a short table depending on $m…","keywords":["complex-valued harmonic functions","zeros","harmonic polynomials","harmonic trinomials","real-variable techniques","roots of unity","zero counting","monotone zero count"],"falsifier":"In an omitted case, say $m\\equiv 2 \\pmod 4$ with $k>0$, enumerate the $2m$ values of $j$ and count how many even and odd $j$ have $\\cos(kj\\pi/m)>0$, $<0$, or $=0$; if the split differs from what the table requires, Theorem 1 fails. Alternatively, for small $m$ and $k$ in such a case, numerically count the zeros of $p(z)=z^m+c(z^k+\\bar z^k)-1$ for $c$ below and above the threshold $c_0$ of Lemma 7 and check that the count is $m$ and $m+N$ respectively.","tokens_in":10148,"feed_emoji":"📊","tokens_out":13768,"duration_ms":121365,"temperature":0.7,"pith_summary":"The paper proves that two one-parameter families of complex-valued harmonic functions have zero counts that change in a completely predictable way as the positive parameter $c$ increases. For $p(z)=z^m+c(z^k+\\bar z^k)-1$ with $\\gcd(m,k)=1$ and $m>|k|$, every zero lies on one of $2m$ rays, and on each ray the problem becomes a real-variable root count. The authors count the rays of each type using roots of unity and obtain Theorems 1 and 2: for $k>0$ the number of zeros increases monotonically from $m$ to $m+N$, and for $k<0$ it decreases monotonically from $M+N$ to $M$, where $N$ and $M$ are determined by $m \\bmod 4$ and the parity of $k$. If these theorems are right, the zero count of these functions is fixed by the exponents and by whether $c$ lies below or above a threshold, with no further dependence on the coefficient. The written proof works through one representative residue class in each theorem and states that the other classes follow by the same counting.","feed_headline":"Harmonic zeros rise and fall by a rule set by m mod 4","feed_subtitle":"As c grows, zeros climb from m to m+N; for k<0 they descend from M+N to M.","key_machinery":"The carrying object is the ray-restricted real function $f_j(r)=(-1)^j r^m + 2c\\cos(kj\\pi/m)r^k -1$, whose positive zeros are exactly the moduli of the zeros of $p$ on the ray at angle $j\\pi/m$. The argument splits into six cases according to the parity of $j$, the sign of $k$, and the sign of $\\alpha=\\cos(kj\\pi/m)$; Lemmas 2 through 11 give 0, 1, or 2 positive roots for each case, with the two-root cases changing at an explicit threshold $c_0$. The remaining work is combinatorial: counting how many of the $2m$ values of $j$ fall into each case, using the fact that $\\cos(kj\\pi/m)=\\operatorname{Re}(\\omega^j)$ for $\\omega=e^{ik\\pi/m}$ and elementary counts of roots of unity with positive, negative, or zero real part.","core_discovery":"The central discovery is a pair of exact zero-counting theorems for the harmonic functions $p(z)=z^m+c(z^k+\\bar z^k)-1$. Theorem 1, for $k>0$ with $\\gcd(m,k)=1$, says that as $c$ increases through the positive reals the number of zeros grows monotonically from $m$ to $m+N$, where $N=m$ for $m\\equiv 0 \\pmod 4$, $N=m-1$ or $m+1$ for $m\\equiv 1 \\pmod 4$ according as $k$ is odd or even, $N=m-2$ for $m\\equiv 2 \\pmod 4$, and $N=m+1$ or $m-1$ for $m\\equiv 3 \\pmod 4$ according as $k$ is odd or even. Theorem 2, for $k<0$, says the zero count decreases monotonically from $M+N$ to $M$, with paired values $(M,N)=(m+1,m-2)$ for $m\\equiv 0 \\pmod 4$; $(m-1,m+1)$ or $(m,m+1)$ for $m\\equiv 1 \\pmod 4$ according as $k$ is odd or even; $(m-1,m)$ for $m\\equiv 2 \\pmod 4$; and $(m+1,m-1)$ or $(m,m-1)$ for $m\\equiv 3 \\pmod 4$ according as $k$ is odd or even. The proof rests on two structural facts: the imaginary part $r^m\\sin(m\\theta)$ vanishes only on the rays $\\theta=j\\pi/m$, and on the $j$-th ray the modulus $r$ of a zero must satisfy $f_j(r)=(-1)^j r^m + 2c\\cos(kj\\pi/m)r^k -1=0$. The six possible forms of $f_j$ have one, two, or no positive real roots, with the two-root cases switching at an explicit threshold, and the root-of-unity count fixes how many rays of each form exist.","pith_inferences":["The same ray-reduction is tied to having exactly one non-real term, so the general $a\\ne b$ case is not accessible by these methods; the paper leaves it open how the maximum zero count there compares, and a numerical survey of small $a,b,m,k$ would be a direct test of whether a similar table exists.","The explicit threshold $c_0$ suggests the radii at which zeros appear or disappear could also be tracked, since the roots of $f_j$ vary continuously in $c$; this would give a quantitative prediction about where in the complex plane new zeros are born.","Because the paper's thresholds $c_0$ are explicit, one can numerically simulate the omitted residue classes, for example $m\\equiv 2 \\pmod 4$, and check the parity splits before relying on the full table; this is a direct computational verification of the unstated cases."],"forward_implications":["For every admissible pair $(m,k)$ with $\\gcd(m,k)=1$, the total number of zeros of $p$ is known exactly at small $c$ and at large $c$: $m$ and $m+N$ for $k>0$, and $M+N$ and $M$ for $k<0$.","The zero count is monotone in $c$, so roots are not created and destroyed repeatedly; each ray's contribution changes only at the threshold where the associated real function gains or loses two roots.","The threshold $c_0$ appearing in the two-root lemmas is explicit in terms of $m$, $k$, and $\\alpha$, so the paper identifies not only the possible counts but the parameter value at which the transition happens.","The residue-class tables in Theorems 1 and 2 cover all $m$ with $m>|k|$ and $\\gcd(m,k)=1$, so the result is complete in the parameter range considered, assuming the omitted cases are as stated."],"supporting_citations":[{"why":"Supplies the closest positive-$k$ harmonic trinomial family and the critical-curve strategy that this paper replaces with real-variable counting.","marker":"[1]"},{"why":"Supplies the negative-$k$ harmonic family with poles whose decreasing zero count motivates Theorem 2.","marker":"[3]"},{"why":"Provides the valence theorem for harmonic polynomials that frames the phenomenon of zero counts exceeding degree.","marker":"[10]"},{"why":"Shows that intermediate zero counts occur in harmonic polynomial families, supporting the monotone transition picture.","marker":"[8]"},{"why":"States the conjecture about maximal zero counts whose proof context motivates bounding zeros of harmonic polynomials.","marker":"[9]"}],"fun_headline_variants":["Harmonic zero counts follow m mod 4 exactly","Exact zero counts for harmonic functions on rays","m mod 4 sets zero counts in harmonic families","New theorems count zeros on harmonic rays exactly","Mod-4 rule dictates harmonic zero numbers"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The theorems assume that the counts of even- and odd-indexed rays with positive, negative, and zero values of $\\cos(kj\\pi/m)$ are exactly as tabulated for every residue class, but the paper writes out the proof for only one class in each theorem and says the others are left to the reader.","fun_headline_variants_meta":{"raw":{"variants":["Harmonic zero counts follow m mod 4 exactly","Exact zero counts for harmonic functions on rays","m mod 4 sets zero counts in harmonic families","New theorems count zeros on harmonic rays exactly","Mod-4 rule dictates harmonic zero numbers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000775,"raw_usage":{"total_tokens":3477,"prompt_tokens":1040,"completion_tokens":2437,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":656,"completion_tokens_details":{"reasoning_tokens":2366}},"tokens_in":656,"tokens_out":2437,"duration_ms":16050,"temperature":1.0,"reasoning_tokens":2366,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:48:15.934398+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"In an omitted case, say $m\\equiv 2 \\pmod 4$ with $k>0$, enumerate the $2m$ values of $j$ and count how many even and odd $j$ have $\\cos(kj\\pi/m)>0$, $<0$, or $=0$; if the split differs from what the table requires, Theorem 1 fails. Alternatively, for small $m$ and $k$ in such a case, numerically count the zeros of $p(z)=z^m+c(z^k+\\bar z^k)-1$ for $c$ below and above the threshold $c_0$ of Lemma 7 and check that the count is $m$ and $m+N$ respectively.","supporting_citations":[{"cited_title":"Zeros of a one-parameter family of harmonic trinomials","cited_arxiv_id":null,"evidence_quote":"Supplies the closest positive-$k$ harmonic trinomial family and the critical-curve strategy that this paper replaces with real-variable counting."},{"cited_title":"Zeros of a family of complex-valued har- monic functions with poles","cited_arxiv_id":null,"evidence_quote":"Supplies the negative-$k$ harmonic family with poles whose decreasing zero count motivates Theorem 2."},{"cited_title":"The valence of harmonic polynomials","cited_arxiv_id":null,"evidence_quote":"Provides the valence theorem for harmonic polynomials that frames the phenomenon of zero counts exceeding degree."},{"cited_title":"On the zeros of polyanalytic polynomials","cited_arxiv_id":null,"evidence_quote":"Shows that intermediate zero counts occur in harmonic polynomial families, supporting the monotone transition picture."},{"cited_title":"Sheil-Small, 02 1992","cited_arxiv_id":null,"evidence_quote":"States the conjecture about maximal zero counts whose proof context motivates bounding zeros of harmonic polynomials."}],"review_version":2}