{"id":"b3f71654-ebda-4205-a753-201111af2391","arxiv_id":"2507.10371","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For stable sequences D(λ,τ) of SU(N) representations, dimensions satisfy dim(D(λ,τ),N) = (-1)^{Area(λ)+Area(τ)} dim(D(τ,λ),-N) and second-order Casimir eigenvalues satisfy C(D(λ,τ),N) = -C(D(λ^T,τ^T),-N).","lead":"A sign-switching trick that relates SU(N) representation theory at N and at -N is extended to a family of representations whose diagrams grow with N, including the adjoint representation. Two exact duality formulas, one for dimensions and one for Casimir eigenvalues, are proved; they tighten the mathematical foundations of the Vogel universality program used in modern Lie theory and knot theory.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proposition 2 rests entirely on the unproven, self-cited Casimir formula (32)-(34), and Proposition 1's hook bookkeeping has unstated steps; both can be settled by direct computation.","rationale":"The reader's weakest_assumption identified exactly the same primary concern—the unproven, self-cited Casimir formula (32)-(34)—and the same secondary concern about hook-content bookkeeping in Proposition 1. My reading of the paper does not reveal a different, more fundamental flaw. The propositions are plausible and the worked Section 3 example supports Proposition 1; Proposition 2's algebra works on the adjoint and simple checks. The paper's main weakness is that its most consequential input, the Casimir formula, is imported without proof, so the central duality claim is conditional on an external source. That source is self-cited, which raises the verification burden. A direct computational test on small N and small λ,τ would settle whether the concern is real. If the test passes, the paper's claims stand; if it fails, Proposition 2 and the universal-decomposition consequences need revision. Since the reader already issued a CONDITIONAL verdict requiring exactly this verification, I do not change that verdict; I agree with it and add a concrete way to resolve it.","tokens_in":7521,"tokens_out":13099,"duration_ms":150800,"concrete_test":"For small explicit cases, e.g. N=5,7 and λ,τ chosen as (2,1), (1,0,1), (2,0,2) with zeros inserted per (12), construct the actual N-dependent Young diagram from the Dynkin labels. First, compute dim(D(λ,τ),N) by the standard Weyl dimension formula on that diagram and compare both sides of (24), evaluating the right-hand polynomial at −N. Second, compute the second-order Casimir eigenvalue on the same highest weight using the standard (λ,λ+2ρ) formula or an explicit Freudenthal computation with long roots squared 2, and compare with (32)-(34) and with (31). If all small cases agree, the imported formula and hook bookkeeping are corroborated; if any discrepancy appears, the failing term can be located and the duality claim adjusted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim has two linked pillars. Proposition 1's proof in Section 4 is not fully documented: the identification of the b-part with dim(τ^T; su(N−Λ)), and the passage from (26) to (28), invoke hook-content relations that the text leaves as 'evident'. In the worked example, the ranges and arguments of (26)-(28) do not transparently match the Λ-row, T-column rectangle a, so a reader cannot verify the sign and factor structure without re-deriving the whole hook bookkeeping. More importantly, Proposition 2 is a direct corollary of the imported formula (32)-(34), stated as 'given in [15]' without proof, page reference, or independent derivation. Since [15] is the author's own prior paper and the formula is not reproduced with derivation, an error in any term—say the i² subtraction or the 1/N cross-square—would invalidate the Casimir duality (31). The concluding universal-decomposition discussion uses both propositions, so the weakest load-bearing premise is the unverified Casimir formula, with the hook bookkeeping as a secondary fragile step. I do not see an internal contradiction, but I do see an unsupported load-bearing input that the paper itself does not supply.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes an extension of the classical N ↔ −N dimension duality for SU(N) representations to a class of representations whose Young diagrams depend on N, namely stable sequences D(λ, τ) with Dynkin labels (λ_1,...,λ_k,0,...,0,τ_k,...,τ_1). Proposition 1 states a dimension duality dim(D(λ,τ),N) = (−1)^{Area(λ)+Area(τ)} dim(D(τ,λ),−N), and Proposition 2 states an analogous duality for the second-order Casimir eigenvalue at the minimal metric, C(D(λ,τ),N) = −C(D(λ^T,τ^T),−N). The proofs use hook-content bookkeeping for the dimension claim and an imported Casimir formula from the author's earlier preprint [15]. The paper closes with a discussion of consequences for the conjectured universal decomposition of powers of the adjoint representation.","tokens_in":7633,"tokens_out":16272,"duration_ms":183729,"significance":"If the two propositions are correct, the paper gives a nontrivial and natural extension of negative-dimensional duality to an N-dependent family of representations that includes the adjoint, and it provides the corresponding Casimir statement. The dimension duality is supported by a fully worked example, and the Casimir formula (32)-(34) is written in explicit Dynkin-label form, so the claims are directly checkable. The concluding prediction—equal multiplicities for diagrams related by the duality in the universal decomposition of powers of the adjoint—is falsifiable and of interest for Vogel-universality studies. The main weaknesses are that the proof of Proposition 1 is compressed to the point of being incomplete, and Proposition 2 depends entirely on a self-cited formula that is not derived in the present paper.","major_comments":[{"comment":"The proof of Proposition 1 is not self-contained. The identification of the b-part contribution with dim(τ^T; su(N−Λ)) is asserted without exhibiting the hook-content factors or the Z2 argument that produces the equality, and the same holds for the b′/λ contribution. Since this bookkeeping is the core of the dimension duality, the proof needs to be written out explicitly rather than left as 'evident'.","section":"Section 4, Proposition 1"},{"comment":"The hook-denominator formulas for the rectangles a and a′ are stated with index ranges that do not match the declared geometry. The text says a is a rectangle with Λ rows and T columns, yet (26) runs i=1,...,T and j=1,...,Λ; when T>Λ the symbol l_i is undefined, and when Λ>T the symbol t_j is undefined. Moreover, the change of variables i→T−i+1, j→Λ−j+1 used to obtain (28) is not shown, so the sign factor (−1)^{Area(a)} and the equality with (26) cannot be checked from the printed text. These steps must be written out.","section":"Section 4, Eqs. (26)-(28)"},{"comment":"Proposition 2 is a direct corollary of the Casimir formula (32)-(34), which is imported from the author's own preprint [15] without proof or a precise pointer. This formula is the sole load-bearing input for the Casimir duality; if any term in (32)-(34) is incorrect, Eq. (31) fails. The paper should either reproduce a derivation of (32)-(34) in an appendix or state exactly where in [15] the formula is proved.","section":"Section 5, Proposition 2"},{"comment":"The duality map in Proposition 2 is D(λ,τ) → D(λ^T,τ^T), whereas Proposition 1 concerns D(λ,τ) → D(τ,λ) and Eq. (25) states dim(D(λ,τ),N)=dim(D(τ^T,λ^T),N). The paper should clarify how transposition acts on the pair (λ,τ) and explicitly justify why the Casimir statement is written with D(λ^T,τ^T) rather than D(τ^T,λ^T); if the formula (32) is symmetric in λ and τ, this should be stated.","section":"Section 5, Eq. (31) vs. Proposition 1 and Eq. (25)"}],"minor_comments":[{"comment":"There are several typos and grammatical errors, including 'repectively', 'transformes', 'finishs', and the malformed display 'N um' in Eq. (4); the manuscript needs a careful copyedit.","section":"Throughout"},{"comment":"The abstract says 'for that representations'; it should read 'for those representations'.","section":"Abstract"},{"comment":"Reference [16] lists the page range as '379-338', which appears to be a typo and should be corrected.","section":"References"},{"comment":"The sums in (52) are written with upper limit N−1, while the Casimir formula (32) uses k; the ranges should be made consistent or the notation should be clarified.","section":"Section 5, Eq. (52)"},{"comment":"The worked example would be easier to verify if the diagrams (13) and (18) were accompanied by explicit row lengths for a concrete small N, so that Area(λ), Area(τ), and the sign (−1)^{Area(λ)+Area(τ)} could be checked without reading the ellipses.","section":"Section 3"}],"recommendation":"major_revision","confidential_remarks":"The main risk is the reliance on the author's own prior preprint [15] for the Casimir formula; I would ask the editor to require either a self-contained derivation or a precise quotation. I do not see an internal contradiction, and the stress-test concern about the Casimir formula does land. With the proof gaps filled, the paper would be a modest but publishable contribution to the negative-dimensional duality literature."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [Colleague],\n\nThe paper gives a real extension of the N↔-N duality for SU(N) to a class of N-dependent Young diagrams (stable sequences D(λ,τ)) that includes the adjoint, where the classical King formula fails. The Casimir duality for this class is new as far as I know, and the proof of the sign-change of the constant term via the a_i/b_i parameterization is a genuine calculation. I checked the transposition identity (45)-(51) and it works; the adjoint limit correctly gives C = 2N.\n\nWhere the paper is soft is in the proof of Proposition 1. The hook bookkeeping in Section 4 is compressed: the identification of the b-part with dim(τ^T; su(N−Λ)) is asserted, and the denominator identities (26)-(28) have index ranges that don't transparently match the Λ×T rectangle. The worked example is convincing, but a reader can't verify the general case without redoing the whole count. That's fixable.\n\nMore importantly, Proposition 2 rests entirely on the Casimir formula (32)-(34), quoted from the author's own prior paper [15] without proof or page reference. The formula looks right in the cases I checked, and the rest of the argument follows cleanly from it, but the paper would be much stronger if it either derived the formula in a few lines or pointed to the exact statement in [15]. As it stands, the load-bearing input is self-cited and not independently verified in this manuscript.\n\nThe concluding remarks about universal decomposition are speculative, but the paper is clear about that. The stable-sequence rule and the Casimir normalization are chosen conventions, and the author says so.\n\nOverall: I think the results are very likely correct and are a useful contribution to the negative-dimensional duality literature. The paper deserves a serious referee, but it needs a revision that fills in the hook bookkeeping and makes the provenance of the Casimir formula explicit.","headline":"Correct and potentially useful extension of N↔-N duality to stable sequences, with a dimension proof that needs tightening and a Casimir result resting on a self-cited formula.","tokens_in":8314,"tokens_out":9677,"would_cite":true,"duration_ms":89013,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["17B10","22E46","22E70"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that SU(N)'s N-to-(-N) duality survives for the N-dependent family D(λ,τ), including the adjoint, with dimensions and Casimir eigenvalues following explicit sign rules.","keywords":["SU(N)","N to -N duality","stable sequences","D(λ,τ) representations","adjoint representation","Young diagrams","Casimir eigenvalues","universal decomposition"],"falsifier":"For a small case such as $\\lambda=(2)$, $\\tau=(1)$ with $N=4$, compute the dimension of $D(\\lambda,\\tau)$ by the hook formula and evaluate the resulting polynomial at $N=-4$; if it differs from $(-1)^{\\mathrm{Area}(\\lambda)+\\mathrm{Area}(\\tau)}\\dim(D(\\tau,\\lambda),-4)$, Proposition 1 fails. Similarly, computing $C(D(\\lambda,\\tau),4)$ from the highest-weight formula and comparing it with $-C(D(\\lambda^T,\\tau^T),-4)$ tests Proposition 2 directly.","tokens_in":7144,"feed_emoji":"🔁","tokens_out":17785,"duration_ms":171762,"temperature":0.7,"pith_summary":"The paper extends the known duality that sends $N$ to $-N$ in SU(N) representation theory from ordinary Young diagrams to a family $D(\\lambda,\\tau)$ of $N$-dependent diagrams, the simplest member being the adjoint representation. It proves that dimensions of these representations transform covariantly under the duality, with a sign fixed by the areas of the two component diagrams, and that the quadratic Casimir eigenvalues change sign under the duality after transposing both diagrams. These identities restore a duality that fails under the literal reading of the classical formula for the adjoint. The stated motivation is universality: the decomposition of powers of the adjoint into Casimir eigenspaces is conjectured to be independent of the algebra's parameters, and the duality predicts equal multiplicities for paired diagrams.","feed_headline":"N to -N duality restored for adjoint-like SU(N) reps","feed_subtitle":"Dimensions and Casimir eigenvalues flip with a sign set by the two Young-diagram areas.","key_machinery":"The machinery is the three-block decomposition of the Young diagram of $D(\\lambda,\\tau)$: a subdiagram $\\lambda$, a rectangle $a$ of height equal to the number of rows of $\\lambda$ and width equal to the number of rows of $\\tau$, and a lower block $b$ shaped by $\\tau$. Under the duality the partner diagram $D(\\tau,\\lambda)$ is assembled from $\\tau$, the transposed rectangle $a'$, and the complement of $\\lambda$, so $b$ and $\\tau$ exchange roles while $a$ turns into $a'$. The proof of the dimension identity is bookkeeping of hook numbers block by block: the $b$-contribution equals the dimension of $\\tau^T$ in a smaller SU($N-\\Lambda$) algebra, matching the $\\tau$-block of the partner. For the Casimir, the load-bearing object is the explicit polynomial (32)--(34) for the eigenvalue at the minimal metric, plus the decomposition of the full eigenvalue into the $\\lambda$-part, the $\\tau$-part, and a cross term $(2/N)\\sum_i i\\lambda_i\\sum_i i\\tau_i$; each piece obeys the expected sign rule separately.","core_discovery":"The central findings are two identities for the stable sequence $D(\\lambda,\\tau)$ of SU(N) representations, whose Dynkin labels (the standard coordinates on the space of highest weights) are $(\\lambda_1,\\dots,\\lambda_k,0,\\dots,0,\\tau_k,\\dots,\\tau_1)$. Proposition 1 states that $\\dim(D(\\lambda,\\tau),N)=(-1)^{\\mathrm{Area}(\\lambda)+\\mathrm{Area}(\\tau)}\\dim(D(\\tau,\\lambda),-N)$, so swapping the two component diagrams and reversing the sign of $N$ changes the dimension only by a sign. Proposition 2 states that the second-order Casimir eigenvalue at the minimal metric satisfies $C(D(\\lambda,\\tau),N)=-C(D(\\lambda^T,\\tau^T),-N)$, with transposed diagrams appearing on the negative-rank side. The proof of Proposition 1 divides the Young diagram into three blocks and matches each block's hook contribution under $N\\leftrightarrow -N$; Proposition 2 follows by applying the transposition sign change to the quadratic terms of the explicit formula (32)--(34) for the eigenvalue, together with the cross term proportional to the product of the areas of $\\lambda$ and $\\tau$.","pith_inferences":["Going beyond the paper: because the proof of Proposition 1 identifies the $b$-block with a dimension of a smaller algebra, the same three-block idea could produce dualities for other rank-dependent families of diagrams, not only those with a single zero-gap in the Dynkin labels.","An unstated corollary of (24) together with the $\\mathbb{Z}_2$ automorphism (25) is a square of four identities relating $D(\\lambda,\\tau)$, $D(\\tau,\\lambda)$, $D(\\lambda^T,\\tau^T)$, and $D(\\tau^T,\\lambda^T)$; tracking the signs would give a purely combinatorial consistency check of the whole system.","Testing the equal-multiplicity prediction on low tensor powers of the adjoint of SU(4) or SU(5) would either strengthen or strain the universality hypothesis; such a check is not performed in the paper itself."],"forward_implications":["The dimension formula applies to every stable sequence $D(\\lambda,\\tau)$, so the adjoint and similar $N$-dependent diagrams now sit inside the $N\\leftrightarrow -N$ duality instead of being exceptions to it.","For $\\tau=0$ the identities reduce to the classical duality (1), so the new statement contains the old one as a limiting case.","In the universal decomposition of powers of the adjoint into Casimir eigenspaces, diagrams exchanged by the duality must have equal multiplicities; this is a concrete, checkable prediction of the universality hypothesis.","When the $\\lambda$ and $\\tau$ diagrams are mutually transposed, the constant term of the Casimir vanishes, a direct analytical consequence of Proposition 2."],"supporting_citations":[{"why":"Supplies the explicit formula (32)--(34) for the Casimir eigenvalue on $D(\\lambda,\\tau)$, from which Proposition 2 is derived by transposition sign changes.","marker":"[15]"},{"why":"Establishes the classical $N\\leftrightarrow -N$ dimension duality for $N$-independent Young diagrams that the paper generalises to stable sequences.","marker":"[2]"},{"why":"Give the standard eigenvalue expression $C=(\\lambda,\\lambda)+2(\\lambda,\\rho)$ used to define the Casimir eigenvalue.","marker":"[12, 13, 14]"},{"why":"Introduce the universal decomposition of powers of the adjoint representation into Casimir eigenspaces, the setting where equal multiplicities are predicted.","marker":"[16, 17]"}],"fun_headline_variants":["Sign flips in SU(N) dimension and Casimir under N to -N","N↔-N duality for stable SU(N) reps with Young diagrams","Adjoint-like SU(N) reps: N to -N sign rule","Area sign controls SU(N) dimension and Casimir flip","Stable reps of SU(N): N to -N duality proven"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The Casimir half of the paper stands on the unproved formula (32)--(34) for the eigenvalue on $D(\\lambda,\\tau)$, taken without proof from the author's earlier work [15]; if that formula is wrong, Proposition 2 cannot be relied on.","fun_headline_variants_meta":{"raw":{"variants":["Sign flips in SU(N) dimension and Casimir under N to -N","N↔-N duality for stable SU(N) reps with Young diagrams","Adjoint-like SU(N) reps: N to -N sign rule","Area sign controls SU(N) dimension and Casimir flip","Stable reps of SU(N): N to -N duality proven"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000153,"raw_usage":{"total_tokens":1158,"prompt_tokens":847,"completion_tokens":311,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":463,"completion_tokens_details":{"reasoning_tokens":230}},"tokens_in":463,"tokens_out":311,"duration_ms":4095,"temperature":1.0,"reasoning_tokens":230,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T17:37:45.194943+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a small case such as $\\lambda=(2)$, $\\tau=(1)$ with $N=4$, compute the dimension of $D(\\lambda,\\tau)$ by the hook formula and evaluate the resulting polynomial at $N=-4$; if it differs from $(-1)^{\\mathrm{Area}(\\lambda)+\\mathrm{Area}(\\tau)}\\dim(D(\\tau,\\lambda),-4)$, Proposition 1 fails. Similarly, computing $C(D(\\lambda,\\tau),4)$ from the highest-weight formula and comparing it with $-C(D(\\lambda^T,\\tau^T),-4)$ tests Proposition 2 directly.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the explicit formula (32)--(34) for the Casimir eigenvalue on $D(\\lambda,\\tau)$, from which Proposition 2 is derived by transposition sign changes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the classical $N\\leftrightarrow -N$ dimension duality for $N$-independent Young diagrams that the paper generalises to stable sequences."}],"review_version":1}