{"id":"1796cc34-29d3-42bb-89fa-33340dd11eea","arxiv_id":"2505.16907","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper gives the first systematic persistent-homology analysis of the Lipschitz landscape on function spaces, proving bounded finite bars for many simply connected targets and an unbounded-bar counterexample for S^3∨S^3.","lead":"This paper studies the set of maps between two shapes, filtered by how much they stretch distances, and records when topological features are born and die using persistent homology. It proves that for many target shapes these features are short-lived, but for a wedge of two 3-spheres some features persist for a long time.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem E's L^{4/3} lower bound rests on the new Lemma 6.5, whose proof is only sketched; if its improved exponent estimate α=9/7 fails, the lower bound collapses.","rationale":"The reader correctly identified the shadowing principle as load-bearing, but the more immediate soft spot is the new internal Lemma 6.5, which is the precise mechanism converting geometric homotopy Lipschitz constants into algebraic dilatation bounds. Theorem E's exponent 4/3 is a direct numerical consequence of Lemma 6.5(ii)'s exponent 9/7; if that exponent is wrong, the theorem's main quantitative claim fails even if the published shadowing theorems are correct. The proof of Lemma 6.5 is not detailed enough for a result of this importance: the construction of φ by 'setting it to the value of the obstruction' is not a rigorous step, and the 'Moreover' endpoint adjustment is asserted without tracking constants. The sign/constant discrepancy in Lemma 8.4's displayed formula strengthens the case that the algebraic estimates need independent verification. This does not amount to a demonstrated falsehood, so the paper should not be rejected; but it should be accepted only conditionally on a complete, checkable proof of Lemma 6.5 (especially the S^3∨S^3 exponent estimate) and a corrected Lemma 8.4 computation.","tokens_in":30632,"tokens_out":31711,"duration_ms":252767,"concrete_test":"Independently re-derive Lemma 6.5(ii) for Y=S^3∨S^3 and X=CP^2×S^3, writing out the induction of Prop. 6.6 degree by degree with explicit norm bounds, and verify that replacing endpoints by ρφ_L, ρψ_L via the O(L^{6/5}) homotopies preserves the bound O((Liph)^{9/7}) with a constant independent of the time length S. Also recompute the η(c2) formula in Lemma 8.4 with correct signs and confirm the coefficient at β(t)=0 is Ω(L^{12}).","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central quantitative claim, Theorem E, is proved by converting a lower bound on algebraic homotopies, Lemma 8.4 (Dil ≥ Ω(L^{12/7})), into a lower bound on geometric homotopies via Lemma 6.5(ii). Specifically, Lemma 8.7 uses Lemma 6.5(ii) to assert that an L'-Lipschitz homotopy between f_L and g_L yields an algebraic homotopy between φ_L and ψ_L of dilatation O((L')^{9/7}); the final exponent L^{4/3} is exactly the quotient (12/7)/(9/7). Lemma 6.5 is new to this paper and its proof is only a sketch: part (i) says one 'sets φ to the value of the obstruction,' which is not a well-defined construction as written, and part (ii)'s 'Moreover' clause, allowing arbitrary endpoint homomorphisms that are homotopic via dilatation O((Liph)^α), is justified only by 'it is clear.' The improved estimate α=9/7 for Y=S^3∨S^3 is stated as a consequence of the degree-7 generators' differential being a product of degree-3 and degree-5 forms, but the induction in Prop. 6.6 has several terms, and the text does not verify that the 'other terms are smaller' uniformly, nor that the time-parametrization and endpoint replacement do not introduce additional powers of Liph. Additionally, the displayed formula for η(c2) in Lemma 8.4 does not satisfy the stated endpoint conditions as written; the correct formula differs by sign and a constant term, and although the Ω(L^{12}) lower bound appears to survive, this signals that the algebraic computations need careful checking. If Lemma 6.5(ii) is valid only with a larger exponent p>9/7, or if the endpoint flexibility fails, then the geometric lower bound would be L^{12/(7p)} < L^{4/3}, and Theorem E's claimed exponent would not follow.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a persistent homology framework for spaces of Lipschitz maps between finite complexes/manifolds, filtered by the Lipschitz constant or its logarithm, and proves a series of theorems about the boundedness of finite bars. The main results are: q-tameness and homotopy invariance of the resulting persistence modules (Theorems 2.6 and 2.8); absence of finite bars for nonpositively curved targets (Theorem A); uniform boundedness of finite bars for based and free loop spaces of simply connected finite complexes (Theorem B, via Sections 4 and 5); uniform boundedness for targets with positive weights or rationall H-space type (Theorems C and D); and, as the central new construction, an example of unbounded finite bars for the simply connected target S^3∨S^3 with domain CP^2×S^3 (and a second domain S^2×S^2×(S^3∨S^3)), with the sharp exponent L^{4/3} (Theorem E, Section 8), together with an optimality theorem showing that no better exponent is possible for 7-dimensional domains (Theorem 9.2).","tokens_in":30900,"tokens_out":17506,"duration_ms":136881,"significance":"If the technical gaps identified below are repaired, this is a substantial contribution to quantitative and persistent homotopy theory. It makes precise Gromov's idea of studying the 'Morse landscape' of the Lipschitz constant on function spaces, and it demonstrates that persistent homology is a meaningful invariant of this landscape, sensitive to rational homotopy type. The paper builds on, and extends, published results of Manin, Berdnikov–Manin, Nabutovsky–Rotman, and Gromov; the framework and the examples are novel. The results are plausible and well-motivated, and the paper is clearly written. However, the proof of the new Lemma 6.5, which is load-bearing for Theorem E and Theorem 9.2, is only a sketch, and a displayed algebraic computation in Lemma 8.4 is incorrect as written. These issues prevent certification of the headline claims in their present form.","major_comments":[{"comment":"Lemma 6.5 is the central new technical tool used in Lemma 8.7 and hence in Theorem E, but its proof is only a sketch and is not a complete argument. In part (i), the step “We then define Φ_{k+1}(V_{k+1})|_{t=1} by setting φ to the value of this obstruction” is not a well-defined construction: the obstruction is a cohomology class in H^{k+1}(X;V_{k+1}), and the text gives no formula for the resulting homomorphism Φ_{k+1} nor a verification that the required factorization through H^*(X) holds. In part (ii), the “Moreover” clause, which is essential because Lemma 8.7 uses the freedom to choose the endpoint homomorphisms η|_{s=0} and η|_{s=S}, is justified only by “From the proof it is clear that the choice … doesn’t affect the end result.” The improved exponent α=9/7 for Y=S^3∨S^3 is asserted after the proof in a single paragraph; the induction in Proposition 6.6 contains several distinct terms (Dil(Φ_k)^{k+2}, Dil(φ)^{k+1}, Dil(ψ)^{k+1}, and isoperimetric constants), and the text does not verify uniformly that the terms other than the product of the degree-3 and degree-5 forms are smaller, nor that the endpoint replacement and time reparametrization do not introduce extra powers of Liph. Since Lemma 6.5(ii) with α=9/7 is precisely the mechanism that converts an L′-Lipschitz geometric homotopy into an algebraic homotopy of dilatation O((L′)^{9/7}) in Lemma 8.7, and since the exponent 4/3 in Theorem E is the quotient (12/7)/(9/7), this gap is load-bearing. A complete proof of Lemma 6.5, including the improved estimate α=9/7, is needed before the claim of Theorem E can be certified.","section":"6.4 (Lemma 6.5)"},{"comment":"In the proof of Lemma 8.4, the displayed formula η(c_2)=yx^2(L^{12} − 1/2 β^2(t)) does not satisfy the endpoint conditions η(c_2)|_{t=0}=η(c_2)|_{t=1}=0 that are required because φ_L(c_2)=ψ_L(c_2)=0. Indeed, with β(0)=−L^6 and β(1)=L^6, this expression evaluates to L^{12}/2 at both endpoints. The explicit homotopy η_L given earlier in the section has η(c_2)=2L^{12} yx^2 t(1−t), which is a different expression that does have the correct endpoints. The Ω(L^{12/7}) lower bound appears to survive once the missing constant term is added (the value at β=0 still contains a term of order L^{12}), but the proof as written is incorrect, and the computation should be redone carefully before the lower bound is relied upon.","section":"8.3 (Lemma 8.4)"}],"minor_comments":[{"comment":"The general statement of Lemma 6.5 yields α=48/35 for Y=S^3∨S^3 and 7-dimensional X, but the proof of Theorem E uses the improved value α=9/7, which is only discussed in the paragraph after the proof. Since this improved value is essential, it should be stated as a separate lemma or corollary with a complete proof.","section":"6.4 (Lemma 6.5)"},{"comment":"The proof for the second family X=S^2×S^2×(S^3∨S^3) consists of “A similar argument to Lemma 8.6 shows…” and “A similar argument to Lemma 8.5 shows…” without spelling out the needed modifications. Given the complexity of the first case, the reader cannot easily verify the second family; the authors should either write out the analogues or clearly state the differences.","section":"8.5"},{"comment":"In the proof of Theorem 7.1, the sentence “For i≥k+1, every i-vector in X is trivial” should read “For i>dim X, every i-vector in X is zero.” Also, the rescaling of the metric on Z by R=Dil(g_c^*m_Y) and the subsequent application of the relative shadowing principle need one more sentence explaining why the relative hypotheses hold for the subcomplex ∂Z×X ∪ Z×A.","section":"7.1 (Theorem 7.1)"},{"comment":"The claim that “any homotopy equivalence between compact Riemannian manifolds can be deformed to a Lipschitz homotopy equivalence” is standard, but it would be helpful to give a proof or a reference, since the corollary depends on it.","section":"2.4 (Corollary 2.10)"},{"comment":"In the discussion of ε-smoothing, the notation M^ε_t is introduced but the dependence on ε of the isometry theorem for q-tame modules (Theorem 2.5) is not discussed; this is a minor clarity issue.","section":"2.3"},{"comment":"The title on the first page contains a spacing artifact (“SP ACES”); this should be corrected in the final version.","section":"Title page"}],"recommendation":"major_revision","confidential_remarks":"The paper is promising and likely correct in outline, but the proof of Lemma 6.5 is not yet a proof, and the incorrect displayed formula in Lemma 8.4 shows that the algebraic computations need careful checking. The authors should be asked to provide a complete proof of Lemma 6.5, including the improved exponent, and to correct Lemma 8.4. If these are fixed, the paper could be acceptable for publication. The reliance on the authors' own published results (Man19, BM22) is not problematic, but the new lemma must stand on its own."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe short version: this is a serious paper that deserves a serious referee, but the headline L^{4/3} example rests on a new lemma whose proof is sketched, and there is a concrete algebraic slip in Lemma 8.4. Neither issue looks fatal, but both need to be checked before the main theorem is trusted.\n\nWhat is new and good: using persistent homology of the Lipschitz filtration on mapping spaces is a natural way to make Gromov's 'Morse landscape' question precise, and the paper gets real theorems out of it. Theorem A (no finite bars for CAT(0) targets) is clean. Theorems B, C, and D—uniform boundedness of finite bars for loop spaces, positive-weight targets, and rational H-spaces—are substantial and appear to follow correctly from the published shadowing principle. Theorem E, the unbounded L^{4/3} example, is exactly the kind of counterexample that makes the theory interesting.\n\nThe soft spot is Lemma 6.5. It does heavy lifting in Section 8, and its proof is a sketch. Part (i) says 'set phi to the value of the obstruction,' which is not a well-defined construction as written. Part (ii)'s 'Moreover' clause, allowing arbitrary endpoints via a homotopy of dilatation O((Liph)^alpha), is justified only by 'it is clear.' The improvement from alpha=48/35 to 9/7 is argued informally: the degree-7 differential is a product of degree-3 and degree-5 forms, but the text does not verify that the other induction terms are uniformly smaller. If the exponent is actually larger than 9/7, the geometric lower bound becomes L^{12/(7p)} with p>9/7, which is weaker than L^{4/3}.\n\nThere is also a concrete error in Lemma 8.4: the displayed formula eta(c2)=yx^2(L^{12}-1/2 beta^2) gives nonzero values at t=0 and t=1, where it should vanish. Replacing it by yx^2(L^{12}-beta^2)/2 fixes the endpoints and still yields the Omega(L^{12/7}) dilatation bound. So the algebra looks repairable, but it signals that the computations need careful verification.\n\nSelf-citation is not a real problem here: the cited theorems of [Man19] and [BM22] are published results with independent proofs. The paper is for topologists, geometers, and TDA-minded readers. It deserves to go to a serious referee, with a clear request to verify Lemma 6.5 and the algebraic computations in Section 8. I would cite it once the details are pinned down.","headline":"A genuinely new quantitative invariant for mapping spaces with impressive first theorems, but the headline L^{4/3} example depends on a sketchy new lemma and an algebraic formula that fails its own endpoint conditions.","tokens_in":31602,"tokens_out":7221,"would_cite":true,"duration_ms":53977,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["55N31","55P62","54C35"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper establishes a homotopy-invariant persistence barcode for the space of Lipschitz maps between manifolds, filtered by log-Lipschitz constant, and shows that for the simply connected target S^3∨S^3 the barcode contains finite bars…","keywords":["persistent homology","Lipschitz constant","mapping space","rational homotopy theory","shadowing principle","barcode","height function","S^3 wedge S^3"],"falsifier":"A concrete disproof would be a sequence of homotopic L_i-Lipschitz maps f_i,g_i: $CP^{2}$×$S^{3}$ → $S^{3}$∨$S^{3}$ together with homotopies whose time-slices all have Lipschitz constant ≤ c $L_i^{{4/3 - ε}}$ for some ε>0 and all i; computing such homotopies explicitly for increasing L would settle whether the 4/3 exponent is true.","tokens_in":30321,"feed_emoji":"🏔️","tokens_out":5525,"duration_ms":41052,"temperature":0.7,"pith_summary":"This paper turns the Lipschitz constant into a filtration of the space of maps between two manifolds, and studies the persistent homology barcode of that filtration. The authors establish that, up to finite interleaving distance, the barcode is a homotopy invariant of the pair, and then ask which targets force finite bars to be bounded in log-Lipschitz length. For targets of nonpositive curvature, for loop spaces, for rationally H-space targets, and for targets with positive rational weights, they prove all finite bars have uniformly bounded length. For the simply connected target $S^{3}$∨$S^{3}$ they show the opposite: homotopic L-Lipschitz maps from $CP^{2}$×$S^{3}$ exist for which every homotopy must pass through maps of Lipschitz constant > $cL^{{4/3}}$, so finite bars grow without bound. The exponent 4/3 is shown sharp.","feed_headline":"Map-space persistence finds forced Lipschitz peaks at L^{4/3}","feed_subtitle":"Finite bars stay bounded for many targets, but S^3∨S^3 forces homotopies through L^{4/3}-steep maps.","key_machinery":"The central mechanism is the shadowing principle of Manin, refined by Berdnikov–Manin for scalable targets: it converts a homomorphism from the Sullivan minimal model of Y to the de Rham algebra of X, together with a formal homotopy of bounded dilatation, into a genuine Lipschitz map or homotopy whose Lipschitz constant is controlled by the dilatation. The paper operates on the algebraic side by constructing explicit homomorphisms φ_L, ψ_L from the minimal model M_Y^* of $S^{3}$∨$S^{3}$ to H^*(X), connecting them by algebraic homotopies of dilatation Ω($L^{{12/7}}$), and then uses Lemma 6.5 to show any real homotopy with time-slice Lipschitz constant C induces such an algebraic homotopy of dilatation O($C^{{9/7}}$), yielding the 4/3 exponent.","core_discovery":"On the paper's own terms, the central claim is that the persistence barcode of (Lip(X,Y), log Lip) is a meaningful quantitative invariant of the pair (X,Y), and that its qualitative features are governed by the rational homotopy type of Y. Theorems C and D give uniform boundedness of the finite bars for targets with positive weights and for rationally H-space targets; Theorem E exhibits the first simply connected counterexample, showing that for Y = $S^{3}$∨$S^{3}$ finite bars of arbitrarily large length occur, and Theorem 9.2 proves a matching upper bound O($L^{{4/3}}$), so the example is sharp.","pith_inferences":["The mechanism behind Theorem E—a quadratic constraint on a homotopy's coefficients that must blow up—suggests that other multi-stage targets, such as wedges of spheres of differing dimensions, will exhibit similar power-law forced peaks, with exponents governed by the degrees of the rational homotopy generators.","One could test the 4/3 bound computationally in low degrees by trying to find explicit homotopies with smaller Lipschitz constant; if one were found, it would disprove the sharpness claim rather than merely refine it.","If the shadowing principle is accepted, these results give a new geometric interpretation of the rational homotopy groups: the degrees of the DGA differential determine how 'expensive' certain homotopies are.","Outside the paper, the boundedness questions for P H_i with i>0 remain open for two-stage targets; the example suggests the answer may hinge on whether the target is scalable."],"forward_implications":["If Theorems C and D hold, then for every finite simplicial complex X and every simply connected target Y with positive weights or rational H-space type, the log-Lipschitz filtered persistence module of Lip(X,Y) has all finite bars uniformly bounded, independent of L.","If Theorem E holds, the barcode of Lip(CP^2×S^3, S^3∨S^3) contains finite bars of arbitrarily long length, giving a quantitative distinction between rationally H-space targets and non-H-space targets.","Sharpness (Theorem 9.2) means the L^{4/3} threshold is intrinsic: for any 7-dimensional domain, homotopic L-Lipschitz maps to S^3∨S^3 are homotopic through O(L^{4/3})-Lipschitz maps, so the unboundedness is tight.","Theorem B gives that for simply connected Y, all finite bars in the based or free Lipschitz loop space have length bounded linearly in homological degree, extending Gromov's quantitative questions about loops."],"supporting_citations":[{"why":"Supplies the shadowing principle that converts algebraic homotopies with bounded dilatation into genuine Lipschitz maps and homotopies with controlled Lipschitz constant.","marker":"[Man19]"},{"why":"Introduces scalable spaces and the improved shadowing principle used to obtain the O(L^{4/3}) upper bound and the sharpness theorem.","marker":"[BM22]"},{"why":"Provides the quantitative Morse-theoretic analysis of loop spaces used for Theorem B and the bounded-length results for based and free loops.","marker":"[NR13]"},{"why":"Lays out the Sullivan minimal model and obstruction theory framework that the algebraic constructions in Sections 6-9 rely on.","marker":"[GM81]"},{"why":"Raises Gromov's quantitative questions about homotopies of Lipschitz maps and supplies methods for loop spaces that the paper extends.","marker":"[Gro99a]"},{"why":"Provides quantitative null-cobordism results, including the version used in Lemma 7.8 for rationally H-space targets.","marker":"[CDMW18]"},{"why":"Gives the isometry theorem and stability of barcodes, justifying the persistent-homology language and interleaving arguments.","marker":"[CSEH07]"}],"fun_headline_variants":["Lipschitz landscape of maps forces L^{4/3} peaks","Persistent homology finds sharp Lipschitz peak L^{4/3}","S^3∨S^3 forces Lipschitz peaks of height L^{4/3}","Map-space barcode: S^3∨S^3 yields unbounded bars","Sharp bound: homotopies forced through L^{4/3}-steep maps"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the shadowing principle gives genuinely controlled Lipschitz maps and homotopies from algebraic data, with constants depending only on dimensions—if that conversion failed in these cases, the uniform boundedness theorems and the $L^{{4/3}}$ lower bound could not be derived.","fun_headline_variants_meta":{"raw":{"variants":["Lipschitz landscape of maps forces L^{4/3} peaks","Persistent homology finds sharp Lipschitz peak L^{4/3}","S^3∨S^3 forces Lipschitz peaks of height L^{4/3}","Map-space barcode: S^3∨S^3 yields unbounded bars","Sharp bound: homotopies forced through L^{4/3}-steep maps"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001098,"raw_usage":{"total_tokens":4505,"prompt_tokens":788,"completion_tokens":3717,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":404,"completion_tokens_details":{"reasoning_tokens":3606}},"tokens_in":404,"tokens_out":3717,"duration_ms":22748,"temperature":1.0,"reasoning_tokens":3606,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T14:54:36.637542+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete disproof would be a sequence of homotopic L_i-Lipschitz maps f_i,g_i: $CP^{2}$×$S^{3}$ → $S^{3}$∨$S^{3}$ together with homotopies whose time-slices all have Lipschitz constant ≤ c $L_i^{{4/3 - ε}}$ for some ε>0 and all i; computing such homotopies explicitly for increasing L would settle whether the 4/3 exponent is true.","supporting_citations":[],"review_version":1}