{"id":"52e48d9e-d963-46aa-ac98-97dd7ac0ebc5","arxiv_id":"1908.09465","paper_version":1,"verdict":"REJECT","confidence":"LOW","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"A weighted version of projective Ricci curvature is defined, with classification statements for Randers and Kropina metrics and for projectively flat metrics with isotropic S-curvature.","lead":"This paper introduces a new curvature quantity in Finsler geometry, the weighted projective Ricci curvature, and derives conditions for Randers and Kropina metrics to be flat with respect to it. Specialists in differential geometry may care if the classifications survive the paper's computational errors.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.4 is false: a non-Randers, non-Kropina locally Minkowski norm (e.g., the L^4 norm) is projectively flat with S=0 and WPRic=0 when F0=F, satisfying all hypotheses yet not Randers or Kropina.","rationale":"I read the paper in good faith and focused on its central claim, Theorem 1.4. The reader's weakest_assumption concerned the expansion (cF + eta)|_s y^s = c_0 F + eta_0 in Section 6, on the grounds that F|_s y^s might not vanish. That specific objection does not land: for any Finsler metric, the spray preserves F because the speed along every geodesic is constant, i.e. y^s F_{x^s} - 2G^s F_{y^s} = 0. Since G^j_s y^s = 2G^j, this is exactly F|_s y^s = 0. Thus the algebraic step in equations (6.4)-(6.5) is justified, and the reader's stated weakest assumption is not the load-bearing problem. However, the theorem is still false. A locally Minkowski norm that is not Randers or Kropina, such as the L^4 norm on R^2, satisfies all hypotheses when F0 = F: it is projectively flat, has S = 0 (hence isotropic S-curvature), and has WPRic0 = 0 (hence isotropic weighted projective Ricci curvature with sigma = 0). Its unit sphere is quartic, while Randers and Kropina unit spheres are quadrics, so it is neither. This counterexample also exposes the incompleteness of the proof: equation (6.5) can degenerate to 0 = 0, and the two cases (sigma = c^2 and sigma != c^2) do not exhaust the possibilities. The paper contains other internal inconsistencies, including the statement of Theorem 1.1 versus its proof, but the direct counterexample to the headline theorem is decisive. For these reasons I recommend REJECT, and I disagree with the reader's identification of the weakest assumption, though I agree with the overall rejection.","tokens_in":14684,"tokens_out":27711,"duration_ms":270146,"concrete_test":"Directly compute the quantities for F = ((y^1)^4 + (y^2)^4)^{1/4} on R^2 with F0 = F. Since G^i = 0, Ric = 0, S = 0, and Sbar = 0, the definition gives WPRic0 = 0, so F satisfies the hypotheses of Theorem 1.4. Then verify that F cannot be written as alpha + beta or alpha^2/beta: the unit sphere of F is the quartic curve (y^1)^4 + (y^2)^4 = 1, whereas both Randers and Kropina unit spheres are quadrics. This computation settles the concern by exhibiting a concrete counterexample to the theorem.","verdict_should_be":"REJECT","load_bearing_attack":"The central claim, Theorem 1.4, is refuted by a locally Minkowski norm. Let F(y) = ((y^1)^4 + (y^2)^4)^{1/4} on R^2. This is a strongly convex Finsler metric, independent of x, so its spray coefficients vanish: G^i = 0. Hence it is projectively flat, its Ricci curvature is Ric = 0, and its S-curvature is S = 0; in particular S is isotropic with c = 0. Choose F0 = F. Then d ln(Sigma) = 0, so Sbar = 0 and WPRic0 = Ric + (n-1)(Sbar^2 + Sbar|_s y^s) = 0. Thus WPRic0 = (n-1) sigma F^2 with sigma = 0, so the metric has isotropic weighted projective Ricci curvature. However, F is neither Randers nor Kropina: a Randers metric F = alpha + beta satisfies the quadratic equation F^2 - 2 beta F - (alpha^2 - beta^2) = 0, so its unit sphere is a quadric, and a Kropina metric F = alpha^2/beta has unit sphere defined by the quadric alpha^2 = beta. The L^4 norm has a quartic unit sphere (y^1)^4 + (y^2)^4 = 1, which cannot be a quadric. The proof of Theorem 1.4 fails exactly in the degenerate case of equation (6.5): with P = 0, c = 0, eta = 0, sigma = 0, equation (6.5) becomes 0 = 0, and the two-case classification does not apply. Thus the conclusion is not merely unproved; it is false as stated.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper defines a weighted projective Ricci curvature WPRic0 = Ric + (n-1)(Sbar^2 + Sbar|_k y^k) with Sbar = (S + d ln Sigma)/(n+1) for a Finsler metric F relative to a fixed Finsler metric F0 and fixed volume forms. It then states four results: a comparison theorem between WPRic0 and Ric under completeness (Theorem 1.1), necessary and sufficient conditions for a Randers metric to be weighted projective Ricci flat (Theorem 1.2), analogous conditions for Kropina metrics (Theorem 1.3), and a classification of projectively flat Finsler metrics with isotropic weighted projective Ricci and isotropic S-curvature as Randers or Kropina metrics (Theorem 1.4). The proofs use standard formulas for sprays, S-curvature, and Ricci curvature, and several examples are included.","tokens_in":15100,"tokens_out":12649,"duration_ms":124058,"significance":"If the results were correct, the paper would provide a useful extension of Shen's projective Ricci curvature and a strong rigidity statement for projectively flat Finsler metrics. The formulas in Theorems 1.2 and 1.3 are concrete and the examples are instructive. However, the central classification Theorem 1.4 is false as stated, and the proof of Theorem 1.1 is internally inconsistent. These are not minor presentation issues; they invalidate the main claims of the paper.","major_comments":[{"comment":"Theorem 1.4 is false as stated. On R^2, let F(y) = ((y^1)^4 + (y^2)^4)^{1/4}, a strongly convex Minkowski norm. Since F is independent of x, its spray coefficients vanish, so F is projectively flat, Ric = 0, and S = 0. Taking F0 = F gives Sigma = 1 and eta = 0, hence Sbar = 0 and WPRic0 = Ric = 0. Thus F has isotropic S-curvature S = (n+1)cF with c = 0 and isotropic weighted projective Ricci WPRic0 = (n-1)sigma F^2 with sigma = 0. But F is neither Randers nor Kropina: a Randers metric F = alpha + beta has unit sphere satisfying a quadratic equation after squaring alpha = 1 - beta, and a Kropina metric F = alpha^2/beta has unit sphere satisfying the quadratic equation alpha^2 = beta, whereas the L^4 norm has the quartic unit sphere (y^1)^4 + (y^2)^4 = 1. In this example equation (6.5) degenerates to 0 = 0, so the case split in the proof does not produce the claimed conclusion.","section":"Theorem 1.4, Section 6"},{"comment":"Independently of the counterexample, the algebraic passage from (6.4) to (6.5) is not justified as written. Expanding the term (cF + eta)|_s y^s requires an identity for F|_s y^s, which the paper neither states nor proves. More concretely, the sign of c0 in (6.5) is inconsistent with the expansion: substituting (cF+eta)^2 + (cF+eta)|_s y^s into (6.4) gives a term -(2c eta + c0)F, not -(2c eta - c0)F, when the equation is rearranged as in (6.5). Even after correcting this sign, the proof gives no argument that every solution of the scalar equation (6.5) must have the Randers or Kropina form, and the degenerate 0 = 0 case is not addressed.","section":"Section 6, Eqs. (6.4)-(6.5)"},{"comment":"Theorem 1.1 and its proof do not agree. The comparison argument proves that phi(0) = 0 for phi = (S + d ln Sigma)/(n+1), i.e. S = -d ln Sigma. This is not the same as S = 0 when F0 is arbitrary. If S = -d ln Sigma, then Sbar = 0 and hence WPRic0 = Ric, so both inequalities WPRic0 >= Ric and WPRic0 <= Ric hold even though S need not vanish. Thus the 'if and only if S = 0' statement in Theorem 1.1 is false unless one additionally assumes d ln Sigma = 0, for example F0 = F with the same volume form. The corollary for PRic requires this additional hypothesis.","section":"Section 3, Theorem 1.1"},{"comment":"The proof of Theorem 1.2 contains an unexplained factor change. Equation (4.4) states S = (n+1)[e00/(2F) - (s0 + rho0)], but equation (4.5) states S = (r00 - 2 alpha s0)/(2F), dropping the factor (n+1). The subsequent computation of S|_m y^m and the final formula (4.7) use the unnormalized expression, so the derivation of Theorem 1.2 does not follow from (4.4) as written. If (4.5) is intended to introduce a normalized S-curvature, this redefinition must be stated explicitly.","section":"Section 4, Eqs. (4.4)-(4.5)"}],"minor_comments":[{"comment":"The sentence 'For simplicity, let us ut eta = ...' contains a typo; it should read 'let us put eta = ...'.","section":"Section 6, first paragraph"},{"comment":"The statement says 'for some scalar function c = c(x)' but the displayed conditions (i) and (ii) do not contain c; this leftover parameter should be removed or explained.","section":"Theorem 1.2, statement"},{"comment":"The symbol beta is used both for a general 1-form in the Randers/Kropina discussion and for the normalized form (1/(n+1)) d ln Sigma in Example 2.1; this reuse is confusing and should be disambiguated.","section":"Section 2, Example 2.1"}],"recommendation":"reject","confidential_remarks":"The main classification theorem is refuted by a simple locally Minkowski example, so I do not see how the central claim can be repaired without substantially changing its scope. The self-citation pattern is noticeable but is not the basis for my recommendation; the mathematical issues are decisive."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper's useful contribution is the definition of WPRic0 as a projective invariant extending Shen's PRic, plus the explicit necessary-and-sufficient conditions for Randers and Kropina metrics to be weighted projective Ricci flat. Those two computational theorems are the kind of thing specialists would want on record, if the algebra checks out. The proof of Theorem 1.1, however, is mis-stated: the argument concludes S = -d ln Sigma, not S = 0, so the theorem should say the weighted S-curvature vanishes. And the Randers section drops a factor (n+1) between (4.4) and (4.5). These look like typos, but they erode confidence in the computations. The real problem is Theorem 1.4. It is false as stated. Take any non-Randers, non-Kropina Minkowski norm on R^n, for instance F = ((y1)^4+(y2)^4)^{1/4} + epsilon sqrt((y1)^2+(y2)^2), epsilon small. This is a regular Finsler metric, independent of x, hence projectively flat with S=0 and Ric=0. With F0=F, WPRic0=0, so it has isotropic weighted projective Ricci and isotropic S-curvature (both with zero function). It is neither Randers nor Kropina. The stress-test's L^4 norm alone is not strongly convex, but the perturbed version fixes that, so the counterexample is genuine. The proof fails in exactly the degenerate case: equation (6.5) becomes 0=0, and the two-case split disappears. Even in the nondegenerate case, the step from a quadratic equation for F to 'Randers' is a non-sequitur: a square root of an arbitrary homogeneous degree-two function is not a Riemannian metric unless it is a quadratic form, and that is never shown. The passage from (6.4) to (6.5) also discards c F|_s y^s without justification. So the main theorem is not just missing a proof; it is wrong. Who gets value from this? Finsler specialists working on projective invariants, and only if they treat Theorems 1.2 and 1.3 as computational claims to be verified. The citation pattern is unremarkable; self-citations are inline. I would not rely on the paper for a theorem. Recommendation: desk reject, or accept only after a major rewrite that removes or sharply restricts Theorem 1.4 and fixes the algebraic slips.","headline":"The new weighted projective Ricci curvature is a reasonable extension, and the Randers/Kropina computations are useful, but the main classification theorem is false: any non-Randers/Kropina locally Minkowski norm satisfies the hypotheses.","tokens_in":15594,"tokens_out":10970,"would_cite":false,"duration_ms":102520,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53B40","53C60"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper introduces weighted projective Ricci curvature and proves that projectively flat Finsler metrics with isotropic weighted projective Ricci and isotropic S-curvature are Randers or Kropina metrics.","keywords":["weighted projective Ricci curvature","Finsler geometry","Randers metric","Kropina metric","S-curvature","projectively flat metric","(α,β)-metric","projective invariance"],"falsifier":"Compute $F_{|s}y^s$ directly for a projectively flat Finsler metric with $G^i=Py^i$: the missing identity is equivalent to $F_{x^s}y^s=2PF$. If one exhibits a projectively flat metric with isotropic $S$-curvature, such as a non-Randers projectively flat $(\\alpha,\\beta)$-metric, for which this equality fails, then equation (6.5) is missing the term $cF_{|s}y^s$ and the paper's argument does not force the metric to be Randers or Kropina.","tokens_in":14515,"feed_emoji":"📐","tokens_out":12246,"duration_ms":107024,"temperature":0.7,"pith_summary":"This paper introduces a new projective invariant, the weighted projective Ricci curvature: for a Finsler metric $F$ and a fixed reference metric $F_0$, it is $\\mathrm{Ric}+(n-1)(\\bar{S}^2+\\bar{S}_{|k}y^k)$, where $\\bar{S}$ packages the $S$-curvature of $F$ with the logarithm of the volume ratio. The paper characterizes exactly when a Randers metric $F=\\alpha+\\beta$ or a Kropina metric $F=\\alpha^2/\\beta$ is weighted projective Ricci flat with respect to its Riemannian part $\\alpha$. Its main theorem says that a projectively flat Finsler metric with isotropic weighted projective Ricci curvature and isotropic $S$-curvature is forced to be one of these two $(\\alpha,\\beta)$-metric types. A sympathetic reader would care because this is a rigidity statement: a mild curvature condition, together with projective flatness, pins the metric down to the two simplest non-Riemannian families.","feed_headline":"Projectively flat Finsler metrics reduce to Randers or Kropina","feed_subtitle":"A new projective invariant forces these flat metrics into the two simplest non-Riemannian families.","key_machinery":"The load-bearing object is the weighted projective Ricci quantity\n$$\\mathrm{WPRic}_0=\\mathrm{Ric}+(n-1)\\left(\\bar{S}^2+\\bar{S}_{|k}y^k\\right),\\qquad \\bar{S}=\\frac{S+d\\ln\\Sigma}{n+1},$$\nwhere $\\Sigma$ is the ratio of the fixed reference volume form to the Busemann-Hausdorff volume form of $F$. Because this combination is invariant under projective changes of metric with a fixed volume form, it pairs naturally with projective flatness. Flatness writes the spray as $G^i=Py^i$, which turns the Ricci trace into $(n-1)(P^2-P_0)$; feeding the two isotropy conditions into that identity produces the quadratic equation whose two solution branches are exactly the Randers and Kropina metric forms.","core_discovery":"The central claim is Theorem 1.4: if a projectively flat Finsler metric has $\\mathrm{WPRic}_0=(n-1)\\sigma F^2$ and $S=(n+1)cF$ for scalar functions $\\sigma(x)$ and $c(x)$, then $F$ is either a Randers metric or a Kropina metric. The proof reduces the two curvature assumptions, together with the projectively flat spray form $G^i=Py^i$ and the resulting Ricci formula $\\mathrm{Ric}=(n-1)(P^2-P_0)$, to a quadratic equation in $F$ whose coefficients depend only on $x$. When the quadratic coefficient $\\sigma-c^2$ vanishes, the solution is a quotient of the form $\\alpha^2/\\beta$, a Kropina metric; otherwise the positive solution has the form $\\sqrt{\\text{quadratic}}+\\text{linear}$, a Randers metric. The paper also proves separate characterizations for Randers and Kropina metrics to be weighted projective Ricci flat with respect to their Riemannian part.","pith_inferences":["If the identity $F_{|s}y^s=0$ used in the proof should fail for some projectively flat metric, the quadratic equation would become higher-degree and the decisive Randers/Kropina dichotomy could admit other metric families; this would leave the classification open rather than false.","The same construction may be tested with other projectively invariant quantities, for instance replacing the $S$-curvature term by the mean Landsberg curvature, to see whether a comparable two-family rigidity appears.","The Funk metric example suggests that on non-complete domains the sign comparison between weighted projective Ricci and Ricci curvature is governed by boundary terms rather than by vanishing $S$-curvature, so the weighted quantity may be useful for probing completeness.","A direct check of $F_{x^s}y^s=2PF$ on known projectively flat families with isotropic $S$-curvature would settle the missing identity by calculation, without needing the full classification."],"forward_implications":["Any projectively flat Finsler metric satisfying the two isotropy conditions is C-reducible, in the sense of belonging to the Randers or Kropina classes, so the rigidity constrains the whole metric function rather than only its geodesic spray.","For a complete Finsler manifold, the inequality $\\mathrm{WPRic}_0\\ge\\mathrm{Ric}$ or $\\mathrm{WPRic}_0\\le\\mathrm{Ric}$ holds if and only if the $S$-curvature vanishes, and then the $S$-curvature is an exact one-form; the Funk metric on the ball shows the completeness assumption cannot be relaxed to positive completeness.","A Randers metric $F=\\alpha+\\beta$ is weighted projective Ricci flat with respect to $\\alpha$ exactly when its Riemannian part has $\\mathrm{Ric}=t^m{}_m\\alpha^2+2t_{00}$ and the divergence $s^m{}_{0;m}=0$.","A Kropina metric $F=\\alpha^2/\\beta$ is weighted projective Ricci flat with respect to $\\alpha$ exactly when equations (1.4) and (1.5) hold, and the proof ties this to the conformal-form condition $r_{00}=\\sigma\\alpha^2$ and hence to vanishing $S$-curvature.","Because weighted projective Ricci curvature is projectively invariant with respect to a fixed volume form, it supplies a curvature quantity that is unchanged under projective equivalence, which is what makes the flatness classification meaningful."],"supporting_citations":[{"why":"It introduces the projective Ricci curvature that $\\mathrm{WPRic}_0$ extends and supplies the $S$-curvature and Randers metric formulas used throughout Sections 4 and 6.","marker":"[14]"},{"why":"It gives the projective Ricci curvature formula and the Randers $\\mathrm{PRic}=0$ classification that Theorem 1.2 adapts to the weighted setting.","marker":"[6]"},{"why":"It supplies the Ricci curvature and $S$-curvature formulas for Kropina metrics used in Theorem 1.3, along with the fact that Ricci-flat Kropina metrics are Berwald.","marker":"[24]"},{"why":"It provides the equivalence among isotropic $S$-curvature, conformal $\\beta$, and $r_{00}=k\\alpha^2$ for Kropina metrics that is used in Remark 5.2 and in the proof of Theorem 1.3.","marker":"[22]"},{"why":"It supplies the Riemann curvature, Ricci, and $S$-curvature background, including the projectively flat spray facts behind equation (6.3).","marker":"[15]"},{"why":"It provides the explicit Randers metric with isotropic $S$-curvature and its Ricci formula used in Example 2.4 to illustrate the weighted projective Ricci expressions.","marker":"[5]"}],"fun_headline_variants":["Weighted projective Ricci forces Finsler metrics into two families","Projectively flat Finsler metrics with weighted Ricci are only two types","New curvature invariant classifies projectively flat Finsler metrics","Rigidity theorem: projectively flat Finsler metrics must be Randers or Kropina","Flat Finsler metrics with weighted Ricci conditions are Randers or Kropina"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof's load-bearing step is an unproved cancellation: when the two isotropy conditions are inserted into the projectively flat Ricci formula, the paper drops a term involving the horizontal derivative of $F$ along $y$, effectively assuming $F_{|s}y^s=0$; the paper gives no reason this derivative must vanish, and if it does not, the Randers/Kropina conclusion does not follow from the stated equations.","fun_headline_variants_meta":{"raw":{"variants":["Weighted projective Ricci forces Finsler metrics into two families","Projectively flat Finsler metrics with weighted Ricci are only two types","New curvature invariant classifies projectively flat Finsler metrics","Rigidity theorem: projectively flat Finsler metrics must be Randers or Kropina","Flat Finsler metrics with weighted Ricci conditions are Randers or Kropina"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000756,"raw_usage":{"total_tokens":3299,"prompt_tokens":824,"completion_tokens":2475,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":440,"completion_tokens_details":{"reasoning_tokens":2375}},"tokens_in":440,"tokens_out":2475,"duration_ms":16398,"temperature":1.0,"reasoning_tokens":2375,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:12:07.401710+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute $F_{|s}y^s$ directly for a projectively flat Finsler metric with $G^i=Py^i$: the missing identity is equivalent to $F_{x^s}y^s=2PF$. If one exhibits a projectively flat metric with isotropic $S$-curvature, such as a non-Randers projectively flat $(\\alpha,\\beta)$-metric, for which this equality fails, then equation (6.5) is missing the term $cF_{|s}y^s$ and the paper's argument does not force the metric to be Randers or Kropina.","supporting_citations":[{"cited_title":"Shen, Differential Geometry of Spray and Finsler Spaces , Kluwer Academic Pub- lishers, Dordrecht, 2001","cited_arxiv_id":null,"evidence_quote":"It introduces the projective Ricci curvature that $\\mathrm{WPRic}_0$ extends and supplies the $S$-curvature and Randers metric formulas used throughout Sections 4 and 6."},{"cited_title":"Cheng, Y","cited_arxiv_id":null,"evidence_quote":"It gives the projective Ricci curvature formula and the Randers $\\mathrm{PRic}=0$ classification that Theorem 1.2 adapts to the weighted setting."},{"cited_title":"Zhang and Y","cited_arxiv_id":null,"evidence_quote":"It supplies the Ricci curvature and $S$-curvature formulas for Kropina metrics used in Theorem 1.3, along with the fact that Ricci-flat Kropina metrics are Berwald."},{"cited_title":"Xia, On Kropina metrics of scalar flag curvature , Differ","cited_arxiv_id":null,"evidence_quote":"It provides the equivalence among isotropic $S$-curvature, conformal $\\beta$, and $r_{00}=k\\alpha^2$ for Kropina metrics that is used in Remark 5.2 and in the proof of Theorem 1.3."},{"cited_title":"Shen, Landsberg Curvature, S-Curvature and Riemann Curvature , MSRI Publi- cation Series, Cambridge University Press, 2004","cited_arxiv_id":null,"evidence_quote":"It supplies the Riemann curvature, Ricci, and $S$-curvature background, including the projectively flat spray facts behind equation (6.3)."},{"cited_title":"Cheng and Z","cited_arxiv_id":null,"evidence_quote":"It provides the explicit Randers metric with isotropic $S$-curvature and its Ricci formula used in Example 2.4 to illustrate the weighted projective Ricci expressions."}],"review_version":1}