{"id":"94d78a6b-f9af-4673-be33-90e348f78349","arxiv_id":"2501.01304","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Non-negatively curved diffusions exhibit total-variation cutoff as soon as the product of spectral gap and mixing time diverges, with a universal mixing-window bound.","lead":"This paper proves that any sequence of Markov diffusions with non-negative Bakry-Emery curvature undergoes the cutoff phenomenon, an abrupt transition to equilibrium, whenever the product of spectral gap and mixing time diverges. Why it matters: it turns a model-by-model analysis into one universal condition and gives an explicit bound on the width of the mixing window in terms of the spectral gap.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 1 is false as stated: for f=1/δ on a µ-set of mass δ<1/e, Ent=log(1/δ), Varent=0, tv=1−δ, so Eq. (10) would give log(1/δ)≤1. The proof of Theorem 1 therefore rests on an invalid premise.","rationale":"The paper's central claim is a universal mixing-window bound for non-negatively curved diffusions. The proof architecture is: Lemma 3 gives entropy decay controlled by varentropy; Lemma 1 reverses Pinsker to lower-bound varentropy by entropy minus a term controlled by tv; Lemma 2 converts the resulting entropy bound into a mixing-time bound. If Lemma 1 is false in the tv regime used, the differential inequality (13) does not follow and the proof of Theorem 1 is unsupported. The counterexample is elementary and independent of the diffusion structure, so it directly tests the stated lemma. The reader already identified Lemma 1 as the weakest premise; our check confirms and strengthens that concern by showing the exact constant 1 fails. This is not a rejection of the scientific claim: a corrected reverse Pinsker inequality of the form Ent ≤ log(1/(1−tv)) + sqrt(Varent)/(1−tv) would still imply cutoff, likely with larger ε-dependent constants. A separate algebraic slip in the optimization step (the inequality log u ≥ 1 − 1/u gives u/(u−1), not t0/(t−t0)) is fixable and does not change the cutoff conclusion, so it is not the main concern. CONDITIONAL is therefore appropriate: the manuscript should be accepted only after Lemma 1 is corrected or its use restricted to a class where it holds, and the constants in Theorem 1 are rechecked.","tokens_in":6484,"tokens_out":25107,"duration_ms":257130,"concrete_test":"Compute both sides of Lemma 1 for the two-level density f = 10 on a measurable set A with µ(A) = 0.1 and f = 0 on A^c: LHS = log(10) ≈ 2.303, RHS = 1 + 0/(1−0.9) = 1. The inequality fails. This single evaluation settles that Eq. (10) cannot be used as a universal lemma; the author should then either quote the correct version (with log(1/(1−tv)) in place of the constant 1) and re-derive the integration in Theorem 1, or supply a structural condition on f_t that restores the claimed bound.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Lemma 1 (Eq. (10)) asserts Ent(X) ≤ 1 + sqrt(Varent(X))/(1−tv(X)) for every random variable X admitting a density. This is not true. Let µ be any probability space with a measurable set A of mass δ ∈ (0, e^{-1}), and set f = 1/δ on A, f = 0 elsewhere. Then Ent(X) = ∫ f log f dµ = log(1/δ), Varent(X) = 0, and tv(X) = 1−δ, so Eq. (10) would imply log(1/δ) ≤ 1, which is false. The failure persists for strictly positive smooth mollifications: take f = η > 0 on A^c and normalize; as η → 0, Ent → log(1/δ), Varent → 0, and tv → 1−δ. This is not a harmless edge case: Theorem 1 applies Eq. (10) at times where tv = 1−ε with ε < 1/2, i.e. precisely the regime 1−tv = ε < 1/2 where the constant 1 is used instead of the true varentropy-free reverse Pinsker bound Ent ≤ log(1/(1−tv)). Consequently, the differential inequality (13) is not justified as written. The main cutoff criterion may still be true with a corrected reverse Pinsker inequality and adjusted constants, but the proof of Theorem 1 in the manuscript is invalid at this step.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a universal bound on the mixing-window width of Markov diffusions with non-negative Bakry-Émery curvature: from any deterministic starting point, wmix(ε) ≤ 3/(λ ε³) + 3√(tmix(1−ε)/(λ ε³)), where λ is the spectral gap. From this it derives a cutoff criterion under the product condition λ tmix(ε) → ∞, covering Riemannian Langevin diffusions and general CD(0,∞) Markov processes. The proof introduces a differential inequality between varentropy and entropy, Lemma 3, and combines it with a reverse Pinsker-type inequality and a mixing-time estimate in terms of entropy; integrating and optimizing the resulting differential inequality yields the window bound. A positive-curvature variant gives the sharper bound wmix(ε) ≤ 3/(κ ε²).","tokens_in":6820,"tokens_out":27116,"duration_ms":255402,"significance":"If the technical flaw described below is corrected, this is a major result: it confirms the product-condition conjecture for the entire CD(0,∞) diffusion class, covers deterministic initial conditions and worst-case starting sets, gives a complete characterization of cutoff on compact manifolds, and unifies several model-specific cutoff proofs. The varentropy differential inequality in Lemma 3 is simple and appears to be new; the integration and optimization steps are transparent and check out. The theorem is parameter-free and the constants are explicit, with no fitted or adjustable quantities. The main obstacle to acceptance is not the scope or importance of the result but the validity of Lemma 1 as stated.","major_comments":[{"comment":"","section":"Section 2, Lemma 1 (Eq. (10))"}],"minor_comments":[{"comment":"","section":"Section 2, Lemma 3"},{"comment":"","section":"Section 2, Lemma 2"},{"comment":"","section":"Section 2, paragraph after Lemma 2"},{"comment":"","section":"Section 1, Corollary 1"}],"recommendation":"major_revision","confidential_remarks":"The paper is by the same author of [19] and [20], and Lemma 1 is reproduced from [19, Lemma 8]. Because the reproduction is false as written, a bare citation is not acceptable here: the corrected lemma must be stated and proved, or quoted exactly with its hypotheses. The rest of the proof appears sound, and I do not see a counterexample to the theorem itself; the flaw is local and repairable. If the corrected lemma is supplied and the constants re-checked, I would support acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: Theorem 1 is probably true, but the proof given here is not valid. Lemma 1, the reverse Pinsker inequality borrowed from your JEMS paper, is false in the form stated, and the proof uses it exactly where it fails.\n\nHere is the counterexample. Take any probability space with a set A of mass δ < 1/e, and set f = 1/δ on A, 0 elsewhere. Then Ent = log(1/δ), Varent = 0, and tv = 1−δ. Inequality (10) would give log(1/δ) ≤ 1, which is false. The worry that this density is not strictly positive does not save the lemma: add a tiny background η on A^c and renormalize. The entropy goes to log(1/δ), the varentropy goes to 0, and the total variation goes to 1−δ, so the inequality still fails for small η. This is not an edge case; the proof of Theorem 1 applies Lemma 1 at times where tv = 1−ε with ε < 1/2, which is precisely the regime of the counterexample.\n\nWhat is genuinely new and good: Lemma 3, the differential inequality d/dt Ent ≤ −Varent/(2t), is original and looks correct under the CD(0,∞) assumption. The strategy of turning varentropy decay into a window estimate is attractive, and the paper is very clearly written. The use of curvature to get local Poincare (12) is standard. Lemma 2 is cited without proof, but that is minor and it is a known result.\n\nThe problem is that Lemma 1 is the bridge that converts Lemma 3 into the differential inequality (13). With Lemma 1 false, the inequality d/dt Ent ≤ −(ε Ent − 1)^2/(2t) is not justified. The proof of the cutoff criterion collapses at this point. My guess is that a corrected reverse Pinsker inequality, for instance Ent ≤ log(1/(1−tv)) + sqrt(Varent)/(1−tv), would restore something close to the stated window bound, and the theorem may be salvageable. But as written, the argument is invalid.\n\nThis is a serious flaw, not a cosmetic one. I would recommend rejecting the current version, but not because the idea is bad. The paper deserves the referee process precisely because the result is important and the repair may be straightforward. I would send it back with a clear request to fix the reverse Pinsker step.","headline":"Elegant theorem, but the proof leans on a reverse Pinsker inequality that is false as stated, so the main argument does not go through.","tokens_in":7329,"tokens_out":10147,"would_cite":false,"duration_ms":99974,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60J60","60J25","58J65"],"pacs":[],"model":"deepseek-v4-flash","headline":"Cutoff holds for every non-negatively curved diffusion under the product condition.","keywords":["cutoff phenomenon","mixing time","spectral gap","Bakry–Émery curvature","varentropy","Langevin diffusion","product condition","diffusion processes"],"falsifier":"Compute, for a concrete non-negatively curved diffusion such as the Ornstein–Uhlenbeck process in high dimension, the quantity $\\frac{d}{dt} \\mathrm{Ent}(X_t) + \\frac{\\mathrm{Varent}(X_t)}{2t}$ along the trajectory; if it is ever positive, Lemma 3 is false, and the bound in Theorem 1 collapses.","tokens_in":6288,"feed_emoji":"📉","tokens_out":5716,"duration_ms":45217,"temperature":0.7,"pith_summary":"This paper resolves the cutoff phenomenon for a broad class of Markov diffusions: any diffusion with non-negative Bakry–Émery curvature, such as a Langevin diffusion in a convex potential, undergoes a sharp transition to equilibrium (cutoff) as soon as its spectral gap times its mixing time diverges. The author proves a universal bound on the width of the mixing window, valid for any deterministic starting point, and derives cutoff from the product condition alone, without model-specific analysis. The proof is elementary and rests on a new differential inequality linking the entropy of the process to its varentropy.","feed_headline":"Cutoff holds for all non-negatively curved diffusions","feed_subtitle":"One spectral-gap condition guarantees the abrupt mixing transition in convex-potential and manifold diffusions.","key_machinery":"The engine of the proof is a differential inequality between the entropy and the varentropy of the process: $\\frac{d}{dt} \\mathrm{Ent}(X_t) \\leq - \\frac{\\mathrm{Varent}(X_t)}{2t}$. It is derived from the local Poincaré inequality $\\mathrm{Var}[g(X_t)] \\leq 2t\\, \\mathbb{E}[\\Gamma g(X_t)]$, which follows from non-negative curvature, together with the chain rule for the carré du champ. The reverse Pinsker inequality $\\mathrm{Ent}(X) \\leq 1 + \\frac{\\sqrt{\\mathrm{Varent}(X)}}{1 - \\mathrm{tv}(X)}$ converts this into an integrable differential inequality for the entropy after the mixing time, whose integration yields the window bound. The positive curvature case replaces the local Poincaré inequality with a time-uniform version.","core_discovery":"On the paper's own terms, the central discovery is that for any non-negatively curved diffusion starting from a deterministic point, the width of the mixing window satisfies $w_{\\mathrm{mix}}(\\varepsilon) \\leq \\frac{3}{\\lambda \\varepsilon^3} + 3 \\sqrt{\\frac{t_{\\mathrm{mix}}(1-\\varepsilon)}{\\lambda \\varepsilon^3}}$ for every $\\varepsilon \\in (0,1/2)$, where $\\lambda$ is the spectral gap. From this estimate, cutoff follows whenever the product condition $\\lambda\\, t_{\\mathrm{mix}}(\\varepsilon) \\to \\infty$ holds: the ratio $t_{\\mathrm{mix}}(1-\\varepsilon)/t_{\\mathrm{mix}}(\\varepsilon)$ tends to $1$ for every $\\varepsilon \\in (0,1)$. Under the stronger positive curvature condition $\\mathrm{CD}(\\kappa,\\infty)$, the sharper bound $w_{\\mathrm{mix}}(\\varepsilon) \\leq \\frac{3}{\\kappa \\varepsilon^2}$ holds, and cutoff follows when $\\kappa\\, t_{\\mathrm{mix}}(\\varepsilon) \\to \\infty$. The result applies on Euclidean spaces and on weighted Riemannian manifolds, and extends to worst-case mixing times by taking a maximum over initial states.","pith_inferences":["The entropy–varentropy differential inequality is likely to hold for a wider class of Markov processes with appropriate curvature bounds, potentially yielding cutoff criteria for discrete chains that satisfy a comparable local Poincaré inequality.","Because the bound depends only on the spectral gap, it suggests that within this class the mixing window is controlled by a single scalar quantity, masking any finer geometric structure.","A natural testable extension is to sharpen the constant in Theorem 1 for specific models, such as the Ornstein–Uhlenbeck process, where the exact window may be computed and compared to the bound."],"forward_implications":["Any sequence of Langevin diffusions in convex potentials with $\\lambda\\, t_{\\mathrm{mix}}(\\varepsilon) \\to \\infty$ exhibits cutoff, removing the need for model-specific mixing-time analyses.","On compact manifolds, cutoff for non-negatively curved diffusions is completely characterized by the product condition, since the condition is necessary there.","The bound extends to worst-case mixing times by maximizing over initial states, so the cutoff criterion holds uniformly over compact state spaces.","Under positive curvature, the mixing window is at most $\\frac{3}{\\kappa \\varepsilon^2}$, giving a sharper quantitative statement whenever a positive curvature lower bound is available."],"supporting_citations":[{"why":"Supplies the reverse Pinsker inequality (Lemma 1) and the spectral-gap mixing-time estimate (Lemma 2) that are combined with the new differential inequality.","marker":"[19]"},{"why":"Provides the Bakry–Émery calculus, including the sub-commutation condition and its consequences such as local Poincaré inequalities.","marker":"[3]"},{"why":"Standard reference for curvature conditions and the implied local Poincaré inequality used in Lemma 3.","marker":"[4]"},{"why":"The source of the product condition conjecture that this paper confirms for non-negatively curved diffusions.","marker":"[17]"},{"why":"Shows the product condition fails to imply cutoff for general reversible processes, highlighting the role of the curvature assumption.","marker":"[10]"}],"fun_headline_variants":["Cutoff proven for every non-negative curvature diffusion","One gap condition drives all non-negatively curved diffusions to cutoff","Simple relation yields universal cutoff for non-negative curvature","Long-standing mixing cutoff problem resolved for non-negative curvature","Abrupt mixing transition proven for all non-negative curvature"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof rests on the reverse Pinsker inequality (Lemma 1), borrowed from reference [19] and not proved in the paper, which is what turns the decay of varentropy into a usable differential inequality for the entropy; if that inequality failed for these diffusions, the window bound would not follow.","fun_headline_variants_meta":{"raw":{"variants":["Cutoff proven for every non-negative curvature diffusion","One gap condition drives all non-negatively curved diffusions to cutoff","Simple relation yields universal cutoff for non-negative curvature","Long-standing mixing cutoff problem resolved for non-negative curvature","Abrupt mixing transition proven for all non-negative curvature"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001239,"raw_usage":{"total_tokens":5065,"prompt_tokens":907,"completion_tokens":4158,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":523,"completion_tokens_details":{"reasoning_tokens":4081}},"tokens_in":523,"tokens_out":4158,"duration_ms":31993,"temperature":1.0,"reasoning_tokens":4081,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:31:15.616278+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute, for a concrete non-negatively curved diffusion such as the Ornstein–Uhlenbeck process in high dimension, the quantity $\\frac{d}{dt} \\mathrm{Ent}(X_t) + \\frac{\\mathrm{Varent}(X_t)}{2t}$ along the trajectory; if it is ever positive, Lemma 3 is false, and the bound in Theorem 1 collapses.","supporting_citations":[{"cited_title":"The cutoff phenomen on for ergodic Markov processes","cited_arxiv_id":null,"evidence_quote":"Shows the product condition fails to imply cutoff for general reversible processes, highlighting the role of the curvature assumption."}],"review_version":1}