{"id":"8e39e70f-6936-4cb9-ba59-a343d7f6593d","arxiv_id":"2509.18618","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":2.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Survey of isoperimetric problems under lower Ricci curvature bounds, presenting sharp concavity inequalities, recent existence results, and open questions.","lead":"A set of lecture notes on how lower bounds on Ricci curvature shape the isoperimetric profile: the tradeoff between volume and perimeter. The notes survey classical and very recent results, with many proofs sketched and exercises for the reader.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Non-compact case of Theorem 5 depends on the sharp Laplacian comparison (39), whose proof in §4.2.1 is only a sketch; the Step 2 passage from the adimensional bound (42) to (39) is not fully demonstrated.","rationale":"The reader and I identify the same load-bearing concern: the non-compact case of Theorem 5 depends on Theorem 11, and the proof of the key estimate (39) is only a sketch. The note is explicit in its introduction that many arguments are sketched and that it is a motivating guide rather than a complete-proof paper. It also attributes Theorem 11 to [13], after [17,16], so there is real external support for the statement. I found no internal inconsistency in the compact cases: the barrier argument in Case 1 is standard, and the singular-set handling in Case 2 is plausible and consistent with the stated perimeter estimates (13)-(14). The non-compact proof is precisely where the newest and least elementary input is used, and the Step 2 derivation of (39) from (42) is the weakest link: the note does not display the passage from the regular/singular Laplacian decomposition to the sharp Riccati bound. If that passage fails or needs an extra hypothesis, the affine-tangent proof of concavity of I^{n/(n-1)} would not go through. That justifies a conditional verdict, as the reader proposed. I therefore recommend no change to the reader's verdict.","tokens_in":33839,"tokens_out":5929,"duration_ms":49471,"concrete_test":"Independently re-derive (39) from the ingredients in §4.2.1 without importing [17, Theorem 3.3]. Specifically: (i) show from (42) and the decomposition (62)-(63) that (log h_α)' ≤ H holds q-a.e. on the needles, including explicit control of the singular part (Δf)_sing; (ii) set v(r)=h_α(r)^{1/(n-1)} and verify that v''≤0, v(0)=1, v'(0)≤H/(n-1), and then apply the Riccati comparison to get (n-1)v'/v ≤ H/(1+Hr/(n-1)); (iii) integrate the resulting bound to confirm that the function in (37) is non-increasing for r≥0 and is maximized at r=0 on all of R. If any step requires an additional hypothesis not stated in Theorem 11 (e.g., RCD regularity of X or extra regularity of E), then the non-compact proof of Theorem 5 is incomplete as written.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The non-compact case of Theorem 5—and hence the proof of Theorem 12—rests on Theorem 11: the function in (37) is maximized at r=0, which yields the affine tangent for I^{n/(n-1)}. The key input is the sharp Laplacian comparison (39): Δd_E ≤ H/(1 + Hd_E/(n-1)) on X\\E. In the note this is not proved; §4.2.1 gives a 'road-map'. Step 1 derives only the adimensional bounds (42), Δd_E ≤ H on X\\E and Δd_E ≥ H on E. Step 2 then invokes the regular/singular decomposition (62)-(63), the localization formula (64), the needle inequalities (65)-(66), and asserts (log h_α)' ≤ H q-a.e. (67), claiming that 'the sharp estimates in (39) then follow from the standard Riccati comparison'. The decisive passage from (42) to (67)—especially control of the singular part (Δf)_sing and the q-a.e. bound on each needle—is left to [43,17]. If (39) held only in the weaker form Δd_E ≤ H, then the monotonicity of (37) would not follow, and the concavity of |E_r|^{1/n} and of I^{n/(n-1)} would not be established by the presented argument. This is a completeness gap rather than a demonstrated error; the cited published results may fill it, but the note itself does not.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This is a lecture-note survey (arXiv:2509.18618) on isoperimetric problems under lower Ricci curvature bounds. The central analytic claim is Theorem 5, the sharp concavity inequality (3) for the isoperimetric profile: -I''_M I_M >= k + (I'_M)^2/(n-1) in the viscosity sense. The compact cases of Theorem 5 are proved in detail by second variation, with a truncation argument when n >= 8. The non-compact case is reduced, via Theorem 10, to isoperimetric sets in non-collapsed Ricci-limit spaces, and then to Theorem 11, whose proof depends on the sharp Laplacian comparison (39), Delta d_E <= H/(1 + H d_E/(n-1)). The note then derives several consequences: Lévy-Gromov and Bishop-Gromov inequalities, a sharp isoperimetric inequality on Ric >= 0 manifolds with Euclidean volume growth (Theorem 12), and applications to the stable Bernstein problem. The Introduction explicitly says that many arguments are only sketched and that the text is intended as a motivating guide rather than a complete research paper.","tokens_in":34282,"tokens_out":7632,"duration_ms":62827,"significance":"If taken as a survey, the note is useful and well-organized: it collects classical and recent results, states precise open questions, and provides exercises. Its strengths include a fairly complete compact-case proof of Theorem 5, the ODE-comparison proofs of Lévy-Gromov and Bishop-Gromov, a second self-contained proof of the sharp isoperimetric inequality (69) via Brunn-Minkowski, and an unusually transparent discussion of which steps are deferred to the literature. The author also credits prior work explicitly. However, the non-compact case of Theorem 5 and the proof of Theorem 12 rest on Theorem 11, especially on the sharp Laplacian comparison (39), whose proof in Section 4.2.1 is only a road-map. Thus the note is not self-contained at its central load-bearing point, although the cited published papers may well supply the missing details.","major_comments":[{"comment":"The decisive Step 2 is not a proof. The text derives only the adimensional bound (42), then asserts the decomposition (62), the representation (63), the needle inequalities (65)-(66), and the q-a.e. bound (67) as following from (63) and (42). The control of the singular part (Delta f)_sing and the passage from the measure-valued inequality (42) to the pointwise/q-a.e. bound on each needle are nontrivial and are simply referred to [43,17]. Since (39) is the input that makes the maximum of (37) at r=0 follow, and hence the concavity of I^{n/(n-1)} in Case 3 of Theorem 5 and the proof of Theorem 12, this is a load-bearing gap in the presented argument. The final 'standard Riccati comparison' is plausible once (65) and (67) are granted, but those are exactly the unproved steps. The note should either fill this gap or clearly state that Theorem 11 and (39) are imported from [13,17] and not pr","section":"§4.2.1, proof of (39)"},{"comment":"The proof of Theorem 5 in the non-compact case is compressed into a short argument after Theorem 11. It relies on Theorem 10 (stated without proof), on the inequality I_X >= I_M (Exercise 11), and on transferring an affine tangent of I^{n/(n-1)} from the limit space X back to M. The last transfer is only sketched in three lines. For a survey this is acceptable if the role of each external theorem is clearly flagged, but as written the reader may mistake the sketch for a complete proof of Theorem 5 in full generality.","section":"§3.1, Case 3 and §4.2, Proof of Case 3"},{"comment":"Theorem 11 also claims that the function in (37) achieves its maximum at r=0 on the whole real line, including negative r. The proof says only 'by exploiting the analogue of (39) inside E' without stating that analogue. Since E is not smooth and the distance to the complement is not a smooth function, this is another nontrivial step. It should be formulated explicitly and either proved or attributed to [13,17].","section":"§4.2, Theorem 11"},{"comment":"The second case of Theorem 12 uses the equality I_X(v)=I_M(v) from Theorem 10 and the inequality AV_R(X) >= AV_R(M) delegated to Exercise 11. This makes the proof dependent on an exercise in a way that is somewhat unusual for a central theorem. If the note is meant to be self-contained, the AV_R comparison should be proved in the text; if not, the dependence should be explicit.","section":"§5.1, Proof of Theorem 12, Case 2"}],"minor_comments":[{"comment":"The definition of the Hausdorff measure H^k(E) uses a double superscript notation 'H^k(E) := sup_{delta>0} H^k_delta(E) := sup ...' that is non-standard and slightly confusing; consider writing H^k_delta first and then H^k(E) = sup_delta H^k_delta(E).","section":"§2.1"},{"comment":"In the proof of Theorem 6, the step 'regularizing Ric, and using classical PDE analysis' is not elaborated; a precise reference or a few lines would help the reader.","section":"§3.3"},{"comment":"The sentence 'H >= 0, otherwise the bound ... degenerates in finite time, contradicting the fact that E is bounded, and the space is non-compact' is cryptic. Please expand the argument or point to the place in [17] where this is justified.","section":"§4.2.1, end of Step 2"},{"comment":"The note would benefit from a table or list stating which results are proved in the text, which are proved only under extra assumptions, and which are quoted from the literature. Many statements are followed by sketches whose status is not always clear from the section heading alone.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"This is a lecture-note survey rather than a standard research article. The main issue is not a demonstrated mathematical error but a mismatch between the proof-like presentation of Theorem 5 and the actual amount of detail supplied for the non-compact case. The author is transparent about the gaps, and the cited papers likely fill them, but the manuscript itself leaves the central load-bearing inequality (39) only sketched. If the venue expects self-contained proofs, this needs major revision; if the venue welcomes expository surveys, the authors should adjust the framing so that Theorem 11 and (39) are explicitly quoted as known results, not proved here."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this is a survey, not a research paper: no new theorems, and the author says so in the introduction. What it does well is present a clean, readable account of the sharp concavity inequality for isoperimetric profiles (Theorem 5), with a genuinely detailed proof in the compact case, a reasonable sketch in the non-compact case, and a nice set of applications: Levy-Gromov, Bishop-Gromov, the sharp Euclidean isoperimetric inequality under Euclidean volume growth, and the recent connections to the stable Bernstein problem. The exposition is honest and the references are carefully given. I believe the survey will be genuinely useful to graduate students and to nonspecialists who want the shape of the recent work. The soft spot is exactly where the stress-test points. The non-compact case of Theorem 5 rests on Theorem 11, and the key input is the sharp Laplacian comparison (39). In section 4.2.1 the paper gives a road-map, not a proof: Step 1 establishes only the adimensional bounds (42), and Step 2 invokes the regular/singular decomposition, localization, and the needle inequalities (65)-(66), then asserts that (39) follows from Riccati comparison. The passage from (42) to the q-a.e. bound (67) is not shown. This is a real completeness gap in the note, but it is flagged: the author explicitly says the full proof is in [17]. I do not read this as an error or a hidden circularity; it is an unproved lemma in a survey that declares itself a motivating guide. The reader's CONDITIONAL verdict is fair. If you need to verify the claimed result itself, you have to go to A-Pa-Po-S [17], where the sharp comparison is stated and proved in the RCD setting. One more observation: the paper leans heavily on the author's own papers for the deepest inputs (Theorems 10, 11, 12, 14). That is normal for a survey and not a flaw, but it does mean the note is not a self-contained proof of the main non-compact statements. Who is it for? A graduate student or a colleague in a neighboring field who wants the statements, the structure of the arguments, and the bibliography. It deserves a serious referee as a survey; I would recommend accepting the paper to review, but the referee should check that the sketch in section 4.2.1 is faithful to [17] and that the attribution of the non-compact Theorem 5 to A-Pa-Po-S is accurate. I would not cite it as a source for the sharp Laplacian comparison; I would cite the original papers.","headline":"A transparent, well-organized survey of recent sharp isoperimetric comparison results under lower Ricci bounds, with the non-compact case of the main theorem honestly sketched rather than fully proved.","tokens_in":808,"tokens_out":795,"would_cite":false,"duration_ms":21368,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53C21","53C20","49Q20"],"pacs":[],"model":"deepseek-v4-flash","headline":"The note argues that a single differential inequality—the sharp concavity of the isoperimetric profile under a Ricci lower bound—organizes much of isoperimetric comparison geometry, and it sketches how this inequality extends from compact m","keywords":["isoperimetric profile","Ricci curvature lower bound","sharp concavity inequality","sphere isoperimetric comparison","volume comparison","non-collapsed Ricci limit spaces","asymptotic volume ratio","stable minimal hypersurfaces"],"falsifier":"Find, on a complete non-compact manifold with Ric ≥ 0 and positive asymptotic volume ratio, a volume v where I_M^{n/(n-1)} is not concave (e.g., the second derivative changes sign), or exhibit an isoperimetric set E in a non-collapsed Ricci-limit space at a point where the viscosity Laplacian of d_E exceeds H/(1 + H d_E/(n-1)); either would refute the chain of reasoning.","tokens_in":33796,"feed_emoji":"📐","tokens_out":5329,"duration_ms":44252,"temperature":0.7,"pith_summary":"The note argues that a single differential inequality—the sharp concavity of the isoperimetric profile under a Ricci lower bound—organizes much of isoperimetric comparison geometry. It states and sketches a proof that on any complete Riemannian manifold with Ricci curvature at least k, the perimeter profile I(v) satisfies -I I'' ≥ k + (I')^2/(n-1) in the viscosity sense. From this one inequality, the sharp comparison with the round sphere and the Euclidean isoperimetric inequality on nonnegatively curved manifolds with positive asymptotic volume ratio follow by ODE comparison and convergence arguments. The technical crux is the noncompact case, where the proof relies on a sharp Laplacian comparison for distance to isoperimetric sets on limit spaces, which is only sketched.","feed_headline":"Ricci bound makes isoperimetric profile sharply concave","feed_subtitle":"The same inequality yields sphere comparison and a sharp Euclidean isoperimetric bound on nonnegative Ricci manifolds.","key_machinery":"The isoperimetric profile I_M(v) = inf{P(E) : |E| = v} and the sharp concavity inequality -I I'' ≥ k + (I')^2/(n-1). For k=0 it says I^{n/(n-1)} is concave; the proof passes to isoperimetric boundaries, uses second variation with Ric(ν,ν) ≥ k, and in noncompact settings uses a sharp Laplacian comparison Δd_E ≤ H/(1 + H d_E/(n-1)) on the exterior of an isoperimetric set in non-collapsed Ricci limit spaces, together with localization into one-dimensional needles.","core_discovery":"The central claim is inequality (3): for a complete Riemannian n-manifold with Ricci curvature at least k, the isoperimetric profile I_M satisfies -I_M I_M'' ≥ k + (I_M')^2/(n-1) in the viscosity sense. When k=0 this is equivalent to concavity of I_M^{n/(n-1)}. The note proves this in the compact case by second-variation of isoperimetric boundaries (with a singular-set perturbation for n≥8) and, for non-compact manifolds, reduces it to a sharp Laplacian comparison for the distance to isoperimetric sets on non-collapsed limit spaces. From this inequality the note derives the sharp isoperimetric inequality P(E) ≥ n(ω_n AVR(M))^{1/n} |E|^{(n-1)/n} on manifolds with Ric≥0 and positive asymptotic","pith_inferences":["If the sharp concavity inequality is taken as a synthetic formulation of lower Ricci bounds, it suggests a possible characterization of Ric ≥ k through isoperimetric profile concavity alone, analogous to curvature-dimension conditions—this is a plausible reformulation the note does not state explicitly.","The equality cases on model spaces hint that generic Ricci ≥ k manifolds have strictly concave isoperimetric profiles in the sense of I^{n/(n-1)}; one could test numerically whether the gap to equality in (3) controls the Gromov–Hausdorff distance to the model space.","The Laplacian comparison (39) is strong enough that, if established in full for non-smooth spaces with lower Ricci bounds, it would extend the Euclidean isoperimetric inequality (69) to that setting; the note states the extension but the proof is the limiting step."],"forward_implications":["For a compact manifold with Ric ≥ k, the isoperimetric profile satisfies -I I'' ≥ k + (I')^2/(n-1); when k=0 this makes I^{n/(n-1)} concave, giving a sharp quantitative control on how perimeter grows with volume.","ODE comparison with the sphere profile yields the sharp isoperimetric comparison with the round sphere and the sharp volume bound |M| ≤ |S^n|, with rigidity if equality holds.","On nonnegatively curved manifolds with Euclidean volume growth, the sharp Euclidean isoperimetric inequality P(E) ≥ n(ω_n AVR)^{1/n} |E|^{(n-1)/n} holds, and equality forces M to be Euclidean space and E a ball.","A spectral generalization gives a weighted Bonnet–Myers theorem under a spectral condition, with consequences for the stable Bernstein problem in low dimensions.","Existence results follow: on surfaces of nonnegative sectional curvature and on nonnegatively curved manifolds with positive asymptotic volume ratio, isoperimetric sets exist for all or all sufficiently large volumes, while explicit examples show small volumes can fail."],"fun_headline_variants":["Ricci bound sharpens Euclidean isoperimetric inequality","Isoperimetric profile concavity from Ricci curvature","Lower Ricci curvature yields sharp isoperimetric constant","Ricci≥0 gives optimal Euclidean isoperimetric bound","Curvature inequality shapes isoperimetric profile"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The noncompact case of the sharp concavity inequality rests on a sharp Laplacian comparison for distance to isoperimetric sets on non-collapsed Ricci-limit spaces (stated as (39)), whose proof is only sketched and mostly delegated to references; if that comparison fails, the concavity and the derived sharp isoperimetric inequality do not follow from the arguments presented.","fun_headline_variants_meta":{"raw":{"variants":["Ricci bound sharpens Euclidean isoperimetric inequality","Isoperimetric profile concavity from Ricci curvature","Lower Ricci curvature yields sharp isoperimetric constant","Ricci≥0 gives optimal Euclidean isoperimetric bound","Curvature inequality shapes isoperimetric profile"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000202,"raw_usage":{"total_tokens":1155,"prompt_tokens":615,"completion_tokens":540,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":359,"completion_tokens_details":{"reasoning_tokens":465}},"tokens_in":359,"tokens_out":540,"duration_ms":5274,"temperature":1.0,"reasoning_tokens":465,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T15:41:10.421847+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find, on a complete non-compact manifold with Ric ≥ 0 and positive asymptotic volume ratio, a volume v where I_M^{n/(n-1)} is not concave (e.g., the second derivative changes sign), or exhibit an isoperimetric set E in a non-collapsed Ricci-limit space at a point where the viscosity Laplacian of d_E exceeds H/(1 + H d_E/(n-1)); either would refute the chain of reasoning.","supporting_citations":[],"review_version":1}