{"id":"59c8488f-0247-4052-85a0-d03132c836a3","arxiv_id":"2501.07494","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Proves that λ3/n < 1/(2√2) - ε for some positive ε, i.e., Nikiforov's unproved strict bound for k=3, via a new graph operation and an analysis of the two smallest eigenvalues.","lead":"This paper proves that the third largest eigenvalue of an n-vertex graph never reaches the upper limit 1/(2√2) ≈ 0.3536 of n predicted by Nikiforov's bound; it must stay strictly below. The proof introduces a novel eigenvector-based graph operation and reduces the extremal case to a structured family of n/2-regular graphs.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 3.7's exclusion of the equality state rests on unverified Desmos-based numerical inequalities in Stage 4; if either check is wrong, the strict gap in Theorem 1.2 collapses.","rationale":"I read the paper in good faith and follow the logical chain: Nikiforov's bound gives c3 <= 1/(2*sqrt(2)); the paper introduces Operation * to restrict minimisers; Section 3 reduces to n/2-regular invariant graphs; Theorem 3.7 is the heart, proving a strict gap for such graphs; Section 4 extends via o(n^2) perturbations. The novelty and much of the structure are plausible, and the paper honestly states its limitations. The single most load-bearing point is the exclusion of the equality case in Theorem 3.7, because the entire strict-gap result depends on it. Within that exclusion, the critical quantitative step is Stage 4's intersection analysis, where the paper invokes Desmos computations for two cubic inequalities. These are not backed by derivations or certificates. If they are true, they likely close the gap; if false or too weak, the central claim fails. I considered the Section 4 reduction (Claim 4.2) as an alternative concern, since it passes from arbitrary minimisers to approximately n/2-regular invariant graphs. However, that step is supported by o(n^2) edge perturbations and Weyl's inequality, and its remaining gaps are less sharp: even if Claim 4.2 needed more detail, the fundamental numerical exclusion in Stage 4 is what makes the strict inequality possible in the first place. The reader's weakest_assumption identifies exactly this Desmos dependency, and I agree. The appropriate verdict remains CONDITIONAL: the argument is credible but not fully rigorous until these numerical checks are either proved analytically or certified by a verification system.","tokens_in":24542,"tokens_out":5777,"duration_ms":56163,"concrete_test":"Run a single certified numerical verification script (e.g., in Sage or Mathematica with exact rational arithmetic on T and S, or interval arithmetic with rigorous error bounds) that checks the two Stage 4 assertions: (i) for T in [1/9,1/4] and S in [T+7/400, (1-(2*sqrt(T)-1)^2)/4] intersected with [0,1/4], the cubic P(-0.7) > 0 and -0.7 lies to the right of the larger negative root of P'(nu) = 3*nu^2 + 2*nu + 4*(T-S); (ii) for T in [0.055,1/8] and S between the stated lower bound and 1/4, P(-sqrt(2)+2*sqrt(T)) >= 0.001 and -sqrt(2)+2*sqrt(T) is past the second turning point. If either check fails, Theorem 3.7's contradiction is unsound; if both pass, the Desmos assertions should be replaced by the certificate.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is Theorem 1.2, whose proof depends entirely on Theorem 3.7 for invariant n/2-regular graphs. The key step in Theorem 3.7 is Stage 4, which excludes the possibility that nu_1 + nu_2 = -sqrt(2). In Intersection IIa, the proof assumes nu_1 <= -0.7, derives a lower bound S >= T + 7/400 from the turning-point condition, and then asserts 'Directly computing with Desmos, we get that as functions of T, we have P(-0.7) is always positive, contradiction.' In Intersection IIb, it assumes nu_1 + nu_2 <= -sqrt(2), derives a lower bound on S, and asserts that P(-sqrt(2)+2*sqrt(T)) 'is bounded below by a positive number (0.001 suffices)' while the upper endpoint is 'non-negative with the only root at T = 1/8.' These are numerical claims about cubic polynomials depending on parameters T, S, with no analytic proof, no interval-arithmetic certificate, and no supplied code. The proof of Theorem 3.7 uses them to force nu_1 > -0.7 (or, in IIb, to force the exact equality state X=Z=1/2, T=1/8, nu_1=nu_2=-sqrt(2)/2), and then a separate argument rules out simultaneous Type 1 and Type 2 equality. If either numerical inequality is false or too coarse, the contradiction that establishes the strict gap fails, and a sequence of invariant n/2-regular graphs could approach -sqrt(2)/2, invalidating Theorem 3.1 and hence Theorem 1.2. The rest of the paper does not provide an alternative proof of these bounds; it merely refers to visual Desmos plots. Since the strict gap is the paper's main novelty, this is the most load-bearing unverified step.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a strengthened upper bound for the third eigenvalue of a graph, namely that c3 is strictly below the classical bound 1/(2√2). The strategy is to study the closely related quantity λ_{n-1}+λ_n and to show a uniform spectral gap: inf(λ_{n-1}+λ_n)/n > −√2/2 + ε. The proof introduces a new graph operation G*, shows that minimising graphs can be assumed invariant under it, derives structural restrictions (ω ≤ 3, χ ≤ 4), then reduces a hypothetical worst-case sequence to n/2-regular invariant graphs. For those graphs the paper develops a lengthy finite-parameter extremal argument with two eigenvector types and rules out the limiting equality case, completing the proof of Theorems 3.1, 1.2, and 1.1.","tokens_in":24796,"tokens_out":8498,"duration_ms":85429,"significance":"If the argument is correct, this is a significant result: it resolves Nikiforov's omitted case k = 3 and gives the first rigorous proof of a uniform gap below 1/(2√2) for c3. The new operation G* is natural and the structural theorems (clique number ≤ 3, chromatic number ≤ 4, and the canonical front/middle/back decomposition) are elegant and appear to be correct. The proof is self-contained in the sense that no numerical constants are fitted and no external computational evidence is needed at the level of the main theorem. However, the paper is not yet a complete proof: several load-bearing inequalities, especially in Stage 4 of Theorem 3.7, are asserted on the basis of ad hoc Desmos computations rather than proved or certified, and the phase analysis in Stage 5 is similarly visual and informal. These gaps are local in the sense that they can likely be repaired with written-out polynomial bounds or interval arithmetic, but as submitted they prevent the central claim from being fully verified.","major_comments":[{"comment":"The exclusion of the equality state ν1+ν2 = −√2 depends on numerical assertions that are never proved or certified. In Intersection IIa the proof states 'Directly computing with Desmos, we get that as functions of T, P(−0.7) is always positive, contradiction', and in Intersection IIb it states that P(−√2+2√T) 'is bounded below by a positive number (0.001 suffices)' while the upper endpoint has 'the only root at T=1/8'. These are claims about cubic polynomials in parameters T and S, but the polynomials are not written out, no analytic proof is given, and no code or interval-arithmetic certificate is supplied. These checks are load-bearing: they are exactly what forces ν1 > −0.7 in IIa and what leaves only X=Z=1/2, T=1/8 in IIb. If either check is false or too coarse, the contradiction in Stage 4 fails and Theorems 3.7, 3.1, and 1.2 are unsupported. This is the central gap and must be closed by a rigorous proof or a machine-checkable certificate.","section":"§3.1, Stage 4 (Intersections IIa and IIb)"},{"comment":"The displayed smoothing inequalities are obtained by an unproved extremal heuristic. The text says that the minimum of Σ d_a x_a under fixed t and A occurs when d_a = n−k−l for the first t/(n−k−l) indices and x_a = x_b for the remaining indices, 'by considering the continuous generalisation'. This is not a proof, and the resulting inequalities are used in Intersections I, II, IIa and IIb to restrict the feasible region. Since the smoothing inequalities are load-bearing for the strict-gap argument, a rigorous derivation of these bounds must be supplied.","section":"§3.1, Stage 2 (Smoothing inequalities)"},{"comment":"The phase analysis is presented largely through visual Desmos plots and informal assertions about hyperbola branches. For example, the proof says 'we claim that smoothing(c) eliminates anything below the lower intersection' and 'the feasible region lies inside the region bounded by the two intersection points of the branches', but no analytic verification of the relevant convexity, monotonicity, and branch-selection facts is provided. These phase claims determine which of Intersections Ia, Ib, IIa, IIb is active, so they are not merely illustrative. The forward reference 'For reasons justified in Stage 5, we require the second root of this cubic' only compounds the problem, since Stage 5 itself is not formal. A complete proof needs explicit inequalities proving the claimed shape of the feasible regions in each phase.","section":"§3.1, Stage 5 and the final casework of Theorem 3.7"},{"comment":"The transition from the equality-state analysis to the asymptotic contradiction relies on the statement that, by continuity, the matrices get arbitrarily close to the equality state. This presupposes a compactness or subsequence argument for the normalized parameters X, Y, Z, T, a, c, and ν as n grows. The later claims do use averages, but they do not directly prove the needed parameter convergence. This is a more localized gap than the numerical checks, but it is still part of the strict-gap argument and should be made precise.","section":"§3.1, final paragraph before Claims 3.8 and 3.9"}],"minor_comments":[{"comment":"The proof of Corollary 2.9 cites Leonida and Li [7], which is an unpublished preprint. If this result is used only as motivation and for examples, this should be stated explicitly; if it is needed in the proof, the argument should be made self-contained or the citation should be to a published source.","section":"Corollary 2.9"},{"comment":"The statement that the neighbours of vertex i are {a_i, a_i+1, ..., b_i} mod n is ambiguous; please specify the circular-interval convention and the intended ranges of a_i and b_i.","section":"Theorem 2.4"},{"comment":"In the sentence 'mean(c) and smoothing(c) intersect until T = 1/8, similarly with mean(c) and smoothing(c)', the second clause should presumably refer to mean(a) and smoothing(a); please correct this typo.","section":"§3.1, Stage 4"},{"comment":"The proof of Theorem 3.7 is very long but has almost no equation numbering; references such as 'smoothing(c)' and 'Extrema' would be much easier to check if the key displayed formulas and inequalities were numbered.","section":"Throughout §3.1"},{"comment":"The phrase 'the complement of Q is in a perfect elimination ordering' is used before its connection to Theorem 2.4 is explained; a short definition or reference would improve readability.","section":"Theorem 3.2"}],"recommendation":"major_revision","confidential_remarks":"I would not accept the paper in its present form because the central strict-gap claim rests on several unverified numerical and visual assertions. The gaps appear repairable: the polynomials in Intersections IIa and IIb can likely be bounded by interval arithmetic or closed-form calculus, and the phase analysis can be turned into explicit inequalities. I therefore recommend major revision rather than rejection. I also suggest that the editor check the status of the companion preprint [7], on which Corollary 2.9 depends."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the short version: this is a real mathematical advance and the proof strategy is interesting, but the key inequality that produces the strict gap is justified by 'Desmos' plots rather than a check a referee can reproduce. Send it to review, but ask for the numerical checks to be made explicit.\n\nWhat's actually new: the paper proves Nikiforov's unpublished claim c_3 < 1/(2√2) − ε_3, by showing inf(λ_{n−1}+λ_n)/n > −√2/2 + ε. The operation ∗ is a nice idea: it forces minimizers to have interval structure, clique number ≤ 3, chromatic number ≤ 4, and — in the regular case — a 2×2 block form. The reduction to n/2-regular invariant graphs in Section 4 is terse but plausible. The appendix result on C_6 blow-ups is solid. The paper is also honest about what it doesn't do: no explicit ε, no improvement over the n/2-regular case, and the conjecture λ_{n−1}+λ_n ≥ −2n/3 is stated as unproved.\n\nThe soft spot is exactly where you'd expect: Theorem 3.7, Stage 4. The whole strict separation theorem rests on excluding the equality case ν1+ν2 = −√2. That exclusion uses two numerical assertions: P(−0.7) > 0 and P(−√2+2√T) ≥ 0.001, where P is a cubic with parameters. The paper says 'directly computing with Desmos' and gives no derivation, no code, no interval-arithmetic certificate. This is load-bearing, not a cosmetic omission. It's likely fixable — these are polynomial inequalities in two parameters, so Sturm sequences or a small Sage/Mathematica script should settle them. But in the current form a referee cannot verify the central step without redoing the computation. The stress-test note is right to flag this.\n\nThe secondary issue is Claim 4.2: the o(n^2) edge-modification argument is only sketched. It looks plausible, and the perturbation idea is standard, but a full proof needs more care with the eigenvalue perturbation.\n\nThere's no circularity, no fitted constants, and no sign of sloppy thinking. The self-citation to Leonida–Li is used for motivation and a corollary, not as a crutch.\n\nWho should read it: spectral extremal graph theorists, and anyone following Nikiforov's c_k conjectures. It deserves a serious referee. I would not desk-reject it. The correct outcome is probably 'major revision' asking for a verifiable proof of the Stage 4 inequalities and a fuller Claim 4.2. Once those are in place, I'd expect it to be accepted.","headline":"A serious proof of a Nikiforov claim, with one load-bearing numerical gap that a referee can ask to be fixed.","tokens_in":25471,"tokens_out":2839,"would_cite":true,"duration_ms":27309,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05C50","05C35"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that the supremum of the third eigenvalue ratio λ3/n over all n-vertex graphs is strictly smaller than the classical threshold 1/(2√2).","keywords":["third eigenvalue","extremal graph eigenvalues","adjacency spectrum","graph operation","eigenvector types","clique and chromatic number","spectral gap"],"falsifier":"Evaluate the polynomial P(ν1)=$ν1^{3}$+$ν1^{2}$+4(T−S)ν1+4(√T(T+S)−S) from Intersection IIa with interval arithmetic over the stated ranges T∈[1/9,1/8], S∈[T+7/400,1/4] and check that P(−0.7)>0; likewise verify P(−√2+2√T)>0.001 for the second polynomial at the prescribed S-endpoints over T∈[0.055,1/8]. A single counterexample to either inequality, or an explicit sequence of n/2-regular invariant graphs with (λ_{n−1}+λ_n)/n approaching −√2/2, would settle the claim either way.","tokens_in":24204,"feed_emoji":"📐","tokens_out":7137,"duration_ms":57489,"temperature":0.7,"pith_summary":"The paper proves a long-claimed strengthening of the extremal bound for the k-th largest adjacency eigenvalue of a graph, in the first open case k=3. It shows there is a fixed gap ε3>0 such that λ3(G)/n < 1/(2√2) − ε3 for every graph G on n vertices, where 1/(2√2) is the bound obtained from arithmetic-mean/quadratic-mean comparison. The proof works indirectly: it studies the sum of the two smallest eigenvalues λ_{n−1}+λ_n, whose infimum is shown to lie strictly above −√2/2, and then converts this into the bound on λ3 via Weyl's inequality. Along the way the paper introduces a new graph operation, the ∗ operation, that restructures any minimising graph into one with clique number at most 3 and chromatic number at most 4, and reduces the hypothetical worst case to a family of n/2-regular invariant graphs whose spectral analysis can be carried out by explicit eigenvector inequalities.","feed_headline":"Third eigenvalue of a graph is locked below the classical limit","feed_subtitle":"The proof uses a new graph operation to rule out the equality case of the AM-QM bound.","key_machinery":"The central object is Operation ∗: given a graph G and an orthonormal pair of eigenvectors x,y for λ_{n−1}(G), λ_n(G), define G* by i∼j iff x_i x_j + y_i y_j < 0. The spectral minimisation principle in Theorem 2.2 shows λ_{n−1}(G*)+λ_n(G*) ≤ λ_{n−1}(G)+λ_n(G), so a minimal graph can be assumed invariant, G=G*. Invariance gives a circular-arc structure (Theorem 2.4), clique number ≤3 and chromatic number ≤4. For the critical n/2-regular invariant case the adjacency matrix splits as [[Q, J−Q],[J−Q, Q]], and the spectrum of G reduces to the spectra of J and 2Q−J; the eigenvectors of 2Q−J are shown to come in two monotone Types (front-increasing-then-decreasing nonnegative, and always-decreasing), from which a series of boundary inequalities defines a feasible region in (a,c,ν)-space. The proof of Theorem 3.7 traces the minimum of ν1+ν2 across phases of these inequalities and rules out ν1+ν2=−√2.","core_discovery":"The central claim is Theorem 1.2: there exists ε>0 such that inf{ (λ_{n−1}(G)+λ_n(G))/n : |V(G)|=n≥3 } > −√2/2 + ε. Consequently, via the inequality λ3+λ_{n−1} ≤ λ2(K_n) = −1, the third eigenvalue ratio is bounded away from 1/(2√2). The proof eliminates the equality case of the elementary AM-QM bound λ_{n−1}^2 + $λ_n^{2}$ ≤ $n^{2}$/4: a sequence of graphs with the sum converging to −√2/2 would force a specific limit state — parameters X=Z=1/2, T=1/8 and eigenvector components a=c=√2/2 — and the paper shows that this state cannot be approached simultaneously by the two structural types of eigenvectors that the invariant graphs admit. The near-equality graphs are then shown to be close, up to o($n^{2}$) edge changes, to n/2-regular graphs invariant under the new operation, so the contradiction transfers to the general problem.","pith_inferences":["The direct numerical checks in Stage 4 (positivity of P(−0.7) and the 0.001 lower bound for P(−√2+2√T)) could be replaced by rigorous interval arithmetic or a computer-certified proof, so those two evaluations are the first place to audit the argument.","The same ∗_k generalisation in Subsection 2.2 may give analogous strengthened bounds for higher k, although the paper notes that for k≥3 there is no canonical ordering of the vectors, which currently blocks that route.","The equality-state analysis suggests that any minimising sequence, if it existed, would concentrate on block-type constructions; one testable extension is to check whether explicit pivalous or circulant blow-up families achieve the −2/3 infimum in the limit rather than −√2/2."],"forward_implications":["There is a constant ε3>0 such that every n-vertex graph satisfies λ3(G)/n < 1/(2√2) − ε3, settling the long-claimed strengthening for k=3.","The infimum of (λ_{n−1}+λ_n)/n over all graphs of order n is bounded below by −√2/2 + ε, so the trivial AM-QM bound is not tight.","The same argument gives c_{−2} < 1/(2√2) − ε3, a strengthened upper bound on the second smallest eigenvalue in absolute value.","Conjecture 5.1, that λ_{n−1}+λ_n ≥ −2n/3, is verified for all graphs on at most 9 vertices and is compatible with the new bound; if true it would imply c3 = 1/3."],"supporting_citations":[{"why":"Supplies the bound c_k ≤ 1/(2√(k−1)) and the equality-case conditions that the proof aims to violate.","marker":"[11]"},{"why":"Introduces the pivalous and H_{a,b} graph families whose invariant structure and large λ3 motivate the operation and the conjecture.","marker":"[7]"},{"why":"Provides the analogous operation for λ1+λ2 on which Operation ∗ is modelled.","marker":"[5]"},{"why":"Establishes the infimum existence for λ_{n−1}+λ_n used in Theorem 1.2 and in the blow-up reduction.","marker":"[9]"},{"why":"Inspires the algebraic inequality framework used in Stage 1 of Theorem 3.7.","marker":"[4]"},{"why":"Gives the degree-sum bounds used in Claim 4.2 to perturb a general graph into n/2-regular form.","marker":"[10]"}],"fun_headline_variants":["Third eigenvalue bound proven via new graph operation","Ruling out equality: third eigenvalue bound strengthened","Third eigenvalue: Nikiforov's missing proof delivered","Graph operation rules out equality case for third eigenvalue"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof that the infimum is strictly above −√2/2 depends on two numerical inequalities in Stage 4 of Theorem 3.7 — that the cubic P is positive at −0.7 and that at −√2+2√T it is bounded below by 0.001 — which are asserted from direct computation rather than proved or machine-verified; if either sign were wrong, the contradiction forcing strict inequality would fail.","fun_headline_variants_meta":{"raw":{"variants":["Third eigenvalue bound proven via new graph operation","Ruling out equality: third eigenvalue bound strengthened","Third eigenvalue: Nikiforov's missing proof delivered","Graph operation rules out equality case for third eigenvalue"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000997,"raw_usage":{"total_tokens":4265,"prompt_tokens":1029,"completion_tokens":3236,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":645,"completion_tokens_details":{"reasoning_tokens":3177}},"tokens_in":645,"tokens_out":3236,"duration_ms":20262,"temperature":1.0,"reasoning_tokens":3177,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:40:28.267955+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the polynomial P(ν1)=$ν1^{3}$+$ν1^{2}$+4(T−S)ν1+4(√T(T+S)−S) from Intersection IIa with interval arithmetic over the stated ranges T∈[1/9,1/8], S∈[T+7/400,1/4] and check that P(−0.7)>0; likewise verify P(−√2+2√T)>0.001 for the second polynomial at the prescribed S-endpoints over T∈[0.055,1/8]. A single counterexample to either inequality, or an explicit sequence of n/2-regular invariant graphs with (λ_{n−1}+λ_n)/n approaching −√2/2, would settle the claim either way.","supporting_citations":[{"cited_title":"Eigenvalue problems of Nordhaus–Gaddum type","cited_arxiv_id":null,"evidence_quote":"Gives the degree-sum bounds used in Claim 4.2 to perturb a general graph into n/2-regular form."},{"cited_title":"Extrema of graph eigenvalues","cited_arxiv_id":null,"evidence_quote":"Supplies the bound c_k ≤ 1/(2√(k−1)) and the equality-case conditions that the proof aims to violate."},{"cited_title":"On graphs with large third eigenvalue, 2025","cited_arxiv_id":null,"evidence_quote":"Introduces the pivalous and H_{a,b} graph families whose invariant structure and large λ3 motivate the operation and the conjecture."},{"cited_title":"On the sum of two largest eigenvalues of a symmetric matrix","cited_arxiv_id":null,"evidence_quote":"Provides the analogous operation for λ1+λ2 on which Operation ∗ is modelled."},{"cited_title":"Linear combinations of graph eigenvalues","cited_arxiv_id":null,"evidence_quote":"Establishes the infimum existence for λ_{n−1}+λ_n used in Theorem 1.2 and in the blow-up reduction."},{"cited_title":"On a conjecture of V","cited_arxiv_id":null,"evidence_quote":"Inspires the algebraic inequality framework used in Stage 1 of Theorem 3.7."}],"review_version":1}