{"id":"7939b5a3-0ba3-4b45-9d2f-ddfc57d6d9d3","arxiv_id":"2504.20803","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Continuation maps give functorial morphisms between Morse fundamental groups, including on different manifolds via grafted trajectories, and the Morse fundamental group of an interpolation function on M×R is a relative fundamental group.","lead":"This paper constructs \"continuation maps\" for the Morse fundamental group, showing how loops built from one Morse function move to loops for another as the function is deformed. It also defines these maps for different manifolds using grafted trajectories and identifies the Morse fundamental group of interpolation-type functions on M×R with a relative fundamental group.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The grafted continuation map (Thm 2.2) is not established: its proof simply asserts that grafted moduli spaces have the same topology and properties as non-grafted ones, without proving that the key boundary stratum is a disk whose boundary is the image of the relation.","rationale":"I read the paper as attempting to define continuation morphisms for the Morse fundamental group, both on a fixed manifold and between different manifolds via grafted trajectories, and to prove functoriality and a relative-group description for interpolation-type functions. The main theorems (2.1, 2.2, 3.1, 3.3, 4.1) are plausible and no direct contradiction is apparent. The load-bearing weak point is the well-definedness of the continuation map on relations, particularly in the grafted case. Theorem 2.2's proof transfers all topological properties from the non-grafted setting by assertion rather than by argument: the boundary stratum Mgr((z,0),{1}) is not shown to be a disk, and its boundary is not shown to equal φ^gr_H(ℓ). Since the grafted continuation map is one of the paper's central new contributions, this gap is significant. The non-grafted Lemma 2.1 is also schematic, but a product/collar-flow argument along the interval [0,1] would likely repair it, so the non-grafted case is less concerning. The reader's weakest assumption identified the same cluster of disk-identification and grafted-transfer issues; my analysis agrees with that assessment. The appropriate verdict remains CONDITIONAL, so I recommend no change to the reader's verdict.","tokens_in":34332,"tokens_out":29961,"duration_ms":308332,"concrete_test":"Verify Theorem 2.2 in the simplest nontrivial grafted setting: take X=Y=S^1, choose f1=f2 a Morse function with one maximum and one minimum, and take H(z)=z^2. Compute the grafted moduli space Mgr,+((z,0),∅) for the index-2 critical point z of f1. Check explicitly whether the boundary stratum Mgr((z,0),{1}) is a 2-disk and whether its boundary word equals φ^gr_H(ℓ) for ℓ=∂Wu(z). If this fails in a case where the non-grafted disk property holds, the 'same topology and properties' assertion in Theorem 2.2 is false; if it holds, the test shows the transfer can be made rigorous but is still absent from the paper.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The continuation morphism is well-defined only if, for every index-2 critical point z of f1, the boundary component M((z,0),M×{1}) (resp. its grafted analogue) is a 2-disk whose boundary is φ_F(ℓ) (resp. φ^gr_H(ℓ)) for ℓ = ∂Wu(z). Lemma 2.1 gives a schematic gluing argument for this in the non-grafted case. The more serious gap is in Theorem 2.2: after defining grafted moduli spaces, the proof states 'Since the moduli spaces involved have the same topology and properties as in the non-grafted case, the rest of the arguments from the proofs of Lemma 2.2 and Theorem 2.1 hold.' This is the only justification that Mgr((z,0),{1}) is a 2-disk and that its boundary is φ^gr_H(ℓ). But grafted trajectories compactify with breaks both before and after the graft, and the graph-transversality condition introduces a stratum structure genuinely different from the non-grafted case. Without a proof of these identifications, Theorem 2.2 — and hence the grafted continuation map between different manifolds, Theorem 1.3, and its corollaries — is not established. The non-grafted Lemma 2.1 may be repairable by a product/collar-flow argument, but the grafted transfer is not justified anywhere in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper defines continuation maps for the Morse fundamental group of a closed manifold. For two Morse-Smale pairs connected by an interpolation pair (F,G) on M×R, it constructs a morphism ϕ_F between the Morse fundamental groups and proves functoriality, first in the same-manifold case (Theorems 1.1, 1.2, 3.1) and then for grafted trajectories between different manifolds (Theorems 1.3–1.5, 2.2, 3.2, 3.3). It also introduces interpolation-type functions on M×R and proves a relative-fundamental-group description (Theorem 4.1), together with corollaries asserting independence of the Morse data and invariance under diffeomorphisms.","tokens_in":34682,"tokens_out":3648,"duration_ms":37584,"significance":"If the announced results are fully established, the paper would give a direct Morse-theoretic proof of invariance of the Morse fundamental group and a new relative description of the fundamental group for interpolation-type functions. The construction is explicit and the paper contains useful examples, notably Example 4.1 and Remark 4.5. The author is also transparent about the folklore isomorphism and does not claim circularly to prove it. However, several load-bearing topological identifications are only sketched or asserted, especially in the grafted setting, and the proof of the relative theorem relies on an unproved flow argument. These gaps prevent the paper from currently supporting the full range of its claims.","major_comments":[{"comment":"The lemma asserts that M_+((z,0),∅) is topologically a 3-disk and that M((z,0),M×{1}) is a 2-disk with the stated boundary decomposition. The proof does not give a rigorous topological argument for these identifications; it relies on a schematic figure, dimension counts, and a cylinder/disk union heuristic. This is load-bearing because Lemma 2.2 and the proof of Theorem 2.1 use the disk identification to conclude that the boundary of M((z,0),M×{1}) is ϕ_F(ℓ), which is exactly what is needed to show that relations map to relations. A complete proof, for example by an explicit collar-flow or product decomposition of the compactification, should be provided.","section":"Section 2.1, Lemma 2.1"},{"comment":"The proof of the grafted continuation morphism is not established. After defining the grafted moduli spaces, the proof states that 'Since the moduli spaces involved have the same topology and properties as in the non-grafted case, the rest of the arguments from the proofs of Lemma 2.2 and Theorem 2.1 hold.' This is an assertion, not a proof. Grafted trajectories compactify with breaks both before and after the graft, and the graph-transversality condition introduces a stratum structure genuinely different from the non-grafted case. In particular, the paper does not prove that M^gr((z,0),{1}) is a 2-disk whose boundary is ϕ_H^gr(ℓ) for ℓ = ∂W^u(z). Since Theorem 2.2 is the basis for Theorems 1.3–1.5 and Corollary 1.2, this gap affects the central claims of the grafted part of the paper.","section":"Section 2.2, Theorem 2.2"},{"comment":"The proof of Theorem 4.1 relies on the vector field X = ∇F̃ + ρ∇h to flow arbitrary paths into tubular neighborhoods of unstable manifolds of index at most 1, and similarly to flow disks into neighborhoods of unstable manifolds of index at most 2, while preserving homotopy classes. No proof is supplied for this flow-and-project statement. Since the manifold is non-compact, convergence of the flow and control at infinity are not automatic. This is the mechanism for both surjectivity and injectivity of the morphism ϕ, so Theorem 4.1 is not fully established without a rigorous argument.","section":"Section 4, proof of Theorem 4.1"},{"comment":"The proof of Proposition 3.1 says that the descent to the quotient is 'analogous to that of Theorem 2.1' after replacing moduli spaces by trajectories at a fixed angle λ. But the angle restriction is not a product situation, and the strata of the compactified moduli spaces at fixed λ require separate verification. Similarly, Proposition 3.3 asserts that the grafted singularities 'are not affected by the graft' and then applies the non-grafted arguments. These statements are not demonstrated, and they are load-bearing for the functoriality theorems in Section 3.","section":"Section 3, Proposition 3.1 and Section 3.2, Proposition 3.3"}],"minor_comments":[{"comment":"The notation '(f1,g2)' and '(f2,g2)' appears to contain typos: the first should presumably be '(f1,g1)' and the second should denote a pair on M2 in Corollary 1.2.","section":"Introduction, Corollary 1.1 and Corollary 1.2"},{"comment":"The condition 'γ(0)∈M×{u}, 0≤u<1' in the displayed definition is followed by a description that also involves broken trajectories; the notation for the space of unbroken trajectories versus its compactification is not consistently distinguished.","section":"Section 2.1, definition of M+((x,0),∅)"},{"comment":"The domain of γ_+ is written as R_+ in the displayed equation, but the subsequent discussion and Remark 2.1 treat γ_+ as defined on [0,v]. This should be clarified.","section":"Section 2.2, Figure 7 and grafted trajectory definition"},{"comment":"The proof refers to ϕ_0 and ϕ_π/2, but the statement of Theorem 3.2 uses ϕ_H0^gr and ϕ_H1^gr; the relation between the two notations is not explained.","section":"Section 3.2, Theorem 3.2 proof"},{"comment":"The equivalence relation (2) is described in words and the reader must infer how 'forming the oriented boundary of the compactification' interacts with rule (3); a precise enumeration of the allowed replacements would improve readability.","section":"Section 4, Definition 4.3"},{"comment":"Several figures are schematic and not all labels are referenced in the text; for instance, the colors used in Figures 4–6 are described in the captions but the connection to the formal moduli-space notation is often implicit.","section":"Figures 2–6"}],"recommendation":"major_revision","confidential_remarks":"The paper contains a promising construction and a clearly stated set of theorems, but the current manuscript is not yet at the level of proof expected for the journal. The most serious gap is the grafted continuation map: Theorem 2.2 is proved by asserting that the grafted moduli spaces behave like the non-grafted ones, and this transfer is not routine. I would recommend major revision, with the expectation that the author provide a genuine proof of Lemma 2.1 and Theorem 2.2, and a rigorous treatment of the flow argument in Theorem 4.1."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Plainly: this is the first continuation map for the Morse fundamental group that I know of, and the relative-group statement for interpolation-type functions is a real extra. Not a trivial repackaging. But the grafted theorem is not proved as written; the proof simply asserts that grafted moduli spaces behave like non-grafted ones. The stress-test note is right to put this first.\n\nWhat is genuinely new: the non-grafted construction in Section 2.1. It turns boundary strata of M+((z,0),∅) into f2-paths via evf2, checks concatenation, and then has to show relators go to relators. The functoriality section is real work: the interpolation square with two angle parameters, the birth-death and breaking analysis, is a serious attempt. Theorem 4.1 and its proof, with the vector-field flow and the surjectivity/injectivity argument, also look coherent. The examples in Section 4 check out. I agree with the reader that there is no circularity: the continuation map is built from interpolation data and does not assume the isomorphism it wants to establish. Citation pattern is fine.\n\nNow the gaps. Lemma 2.1 asserts that M+((z,0),∅) is a 3-disk with the stated boundary decomposition, and that M((z,0),M×{1}) is a 2-disk. The proof is mostly a schematic picture plus dimension counts. This is probably repairable—the R-coordinate gives a collar away from the critical slice—but it is not a proof as written. More serious is Theorem 2.2. After defining the grafted moduli spaces, the proof says: \"Since the moduli spaces involved have the same topology and properties as in the non-grafted case, the rest of the arguments... hold.\" That sentence carries the entire grafted version. You need Mgr((z,0),{1}) to be a disk whose boundary is φ^gr_H(ℓ), and grafted compactifications genuinely differ: breaks can occur before and after the graft, and the graph-transversality condition adds an extra stratum. That transfer is not automatic and is not proved. As a result, Theorem 1.3 and the grafted corollaries are not established as they stand.\n\nWho should read it: someone working on Morse/Floer homotopy, especially flow categories or stable Morse theory. It should not be desk-rejected. A serious referee should see it, and should require a full proof of Lemma 2.1 and, for the grafted case, either a real argument that the boundary stratum is a disk with the right boundary, or a revised claim that only the non-grafted result is proven. If those gaps close, the paper is a solid contribution.","headline":"A genuinely new but under-proved paper: the non-grafted continuation map is plausible and probably repairable, but the grafted version rests on an unproved assertion that grafted moduli spaces behave like non-grafted ones.","tokens_in":35138,"tokens_out":3413,"would_cite":false,"duration_ms":35916,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57R70","57R58","55Q05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Continuation maps make the Morse fundamental group a functor, even across different manifolds.","keywords":["Morse fundamental group","continuation map","Morse-Smale pair","grafted trajectories","relative fundamental group","interpolation function","functoriality","moduli spaces"],"falsifier":"Take a standard Morse-Smale pair on $S^2$ with one index-2 point $z$, use the interpolation $F$ of Section 2, and compute the compactified stratum $M((z,0),M\\times\\{1\\})$ explicitly: if it is not a single 2-disk, or if its boundary word under $ev_{f_2}$ differs from $\\phi_F(\\ell)$, then Lemma 2.1 fails and the proof that relations map to relations collapses. Alternatively, take $M=S^1$ with a Morse function having two minima and one maximum and $h(s)=s^3-\\frac{3}{2}s^2$: Theorem 4.1 predicts $\\pi_1^{\\mathrm{Morse}}(F,(m,-\\infty))$ is trivial, so exhibiting two inequivalent Morse paths based at $(m,-\\infty)$ would disprove injectivity.","tokens_in":34120,"feed_emoji":"🔁","tokens_out":13876,"duration_ms":113479,"temperature":0.7,"pith_summary":"This paper proves that the continuation map of Morse theory can be lifted from homology to the Morse fundamental group. For two Morse-Smale pairs on the same manifold joined by an interpolation pair $(F,G)$, anti-gradient trajectories of $F$ assemble the Morse steps of $f_1$ into Morse paths for $f_2$, giving a morphism between $\\pi_1^{\\mathrm{Morse}}(f_1,\\ast_1)$ and $\\pi_1^{\\mathrm{Morse}}(f_2,\\ast_2)$ that descends to the quotient and is functorial. Using grafted trajectories, the same construction works for Morse data on different manifolds, producing morphisms that depend only on the isotopy class of the map and are isomorphisms for diffeomorphisms. The paper also defines the Morse fundamental group of an interpolation-type function on $M\\times\\mathbb{R}$ and shows it is isomorphic to the relative fundamental group of $\\mathbb{R}\\times M$ with respect to a sublevel set. These are invariance and naturality statements proven entirely inside Morse theory, without passing through the isomorphism to $\\pi_1(M)$.","feed_headline":"Morse fundamental groups gain continuation maps","feed_subtitle":"New morphisms prove invariance purely inside Morse theory, including between different manifolds.","key_machinery":"The central machinery is the positive moduli space $M^+((x,0),\\emptyset)$, the space of anti-gradient trajectories of the interpolation function $F$ that start at a critical point $(x,0)\\in M\\times\\{0\\}$ and flow upward in the $\\mathbb{R}$-direction. For an index-2 critical point $z$ of $f_1$, Lemma 2.1 asserts that this space is a 3-disk whose boundary decomposes into an $f_1$-side, an intermediate cylinder, and an $f_2$-side $M((z,0),M\\times\\{1\\})$, which is itself a 2-disk; Lemma 2.2 identifies the boundary of that disk with the image $\\phi_F(\\ell)$ of the Morse loop $\\ell$ bounding $z$. This disk-to-loop identification carries relators of $\\pi_1^{\\mathrm{Morse}}(f_1)$ into relators of $\\pi_1^{\\mathrm{Morse}}(f_2)$. In the grafted setting the same role is played by the grafted moduli space $M^{\\mathrm{gr}}((z,0),\\emptyset)$, and in Section 4 the evaluation of steps along the flow, projected onto sublevel sets, establishes the isomorphism to the relative fundamental group. The interpolation square with angle parameters $\\lambda$ and $\\alpha$ is the auxiliary device used to prove functoriality.","core_discovery":"The paper's central claim is that the Morse fundamental group is a natural object under continuation data. Given an interpolation pair $(F,G)$ between Morse-Smale pairs $(f_1,g_1)$ and $(f_2,g_2)$ on a closed manifold $M$, the positive unstable moduli spaces $M^+((x,0),\\emptyset)$ define a map on Morse steps whose concatenation gives a morphism $\\phi_F:\\pi_1^{\\mathrm{Morse}}(f_1,\\ast_1)\\to\\pi_1^{\\mathrm{Morse}}(f_2,\\ast_2)$; Theorem 2.1 states that it descends to the relations. In the grafted case, where $f_1$ and $f_2$ live on different manifolds $X$ and $Y$ and a map $H:X\\to Y$ is used to join flow lines, Theorem 2.2 produces a morphism $\\phi_H^{\\mathrm{gr}}$ between the two Morse fundamental groups. Theorems 3.1 and 3.3 make both constructions functorial, and Theorem 4.1 identifies the Morse fundamental group of an interpolation-type function $F$ with the relative fundamental group $\\pi_1(\\mathbb{R}\\times M, F^{-1}(]-\\infty,c])\\cup\\{\\ast'\\},\\ast')$.","pith_inferences":["The same boundary-decomposition argument should carry over to Floer fundamental groups, where continuation maps are usually defined only at chain level; the missing input is a compactness and transversality analogue of Lemma 2.1.","Because the grafted morphism depends only on the isotopy class of $H$, the construction should descend to a purely Morse-theoretic model of the induced map on $\\pi_1$ for any continuous map between manifolds.","Theorem 4.1 makes the Morse fundamental group of an interpolation-type function computable from the topology of a sublevel set, so continuation maps could be compared by computing relative groups rather than by chasing moduli spaces.","A natural test would be to promote the construction to a Morse fundamental groupoid whose objects are minima, including formal minima at infinity; the Section 4 examples suggest the pointed-set formulation hides part of the concatenation structure."],"forward_implications":["Any two Morse-Smale pairs on the same closed manifold connected by an interpolation pair have isomorphic Morse fundamental groups, with the continuation morphism itself giving the isomorphism (Corollary 1.1).","A diffeomorphism between two manifolds induces an isomorphism between their Morse fundamental groups through the grafted continuation morphism (Corollary 1.2).","Grafted continuation morphisms depend only on the isotopy class of the map $H$: an isotopy changes the morphism by conjugation by a Morse path (Theorem 3.2).","For an interpolation-type function $F$ on $M\\times\\mathbb{R}$, the Morse fundamental group is isomorphic to the relative fundamental group of $\\mathbb{R}\\times M$ based at a sublevel set (Theorem 4.1), which yields an alternative proof of surjectivity of the continuation morphism.","The continuation construction is functorial: the triangle formed by three Morse-Smale pairs commutes up to a canonical isomorphism (Theorems 3.1 and 3.3)."],"supporting_citations":[{"why":"Supplies the moduli-space and Morse-step framework for the Morse fundamental group that the paper adapts.","marker":"[3]"},{"why":"Contains the proof of the folklore theorem $\\pi_1^{\\mathrm{Morse}}(f,\\ast)\\cong\\pi_1(M,\\ast)$, the invariance baseline the continuation morphisms must reproduce.","marker":"[5]"},{"why":"Provides the grafted trajectory construction used to define continuation morphisms between different manifolds.","marker":"[6]"},{"why":"Gives the grafted trajectory set-up in a monopole setting that the paper's grafted moduli spaces are modeled on.","marker":"[10]"},{"why":"Supplies the interpolation-square gluing framework used to prove functoriality via the angle parameters $\\lambda$ and $\\alpha$.","marker":"[7]"},{"why":"Provides standard Morse theory background on unstable manifolds and continuation maps used in the construction.","marker":"[2]"},{"why":"Defines the relative fundamental group $\\pi_1(X,A,\\ast)$ used in Theorem 4.1.","marker":"[9]"}],"fun_headline_variants":["Continuation maps link Morse fundamental groups","Functorial continuation for Morse fundamental groups","Grafted trajectories extend Morse continuation maps","Interpolations identify Morse and relative fundamental groups","Morse continuation maps across distinct manifolds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof rests on Lemma 2.1's assertion that $M^+((z,0),\\emptyset)$ is topologically a 3-disk with the stated boundary decomposition, so that the $f_2$-side is a 2-disk whose boundary is $\\phi_F(\\ell)$; the lemma is justified by a schematic figure and dimension counts rather than a full gluing proof, and the grafted version simply asserts that the same topology carries over.","fun_headline_variants_meta":{"raw":{"variants":["Continuation maps link Morse fundamental groups","Functorial continuation for Morse fundamental groups","Grafted trajectories extend Morse continuation maps","Interpolations identify Morse and relative fundamental groups","Morse continuation maps across distinct manifolds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000266,"raw_usage":{"total_tokens":1614,"prompt_tokens":950,"completion_tokens":664,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":566,"completion_tokens_details":{"reasoning_tokens":598}},"tokens_in":566,"tokens_out":664,"duration_ms":6402,"temperature":1.0,"reasoning_tokens":598,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:19:49.732185+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a standard Morse-Smale pair on $S^2$ with one index-2 point $z$, use the interpolation $F$ of Section 2, and compute the compactified stratum $M((z,0),M\\times\\{1\\})$ explicitly: if it is not a single 2-disk, or if its boundary word under $ev_{f_2}$ differs from $\\phi_F(\\ell)$, then Lemma 2.1 fails and the proof that relations map to relations collapses. Alternatively, take $M=S^1$ with a Morse function having two minima and one maximum and $h(s)=s^3-\\frac{3}{2}s^2$: Theorem 4.1 predicts $\\pi_1^{\\mathrm{Morse}}(F,(m,-\\infty))$ is trivial, so exhibiting two inequivalent Morse paths based at $(m,-\\infty)$ would disprove injectivity.","supporting_citations":[{"cited_title":"grafted Morse- Smale","cited_arxiv_id":null,"evidence_quote":"Supplies the moduli-space and Morse-step framework for the Morse fundamental group that the paper adapts."},{"cited_title":"Then it has a unique critical point ats = 0, which is a maximum, and its limits at±∞ are−∞ (in particular, it is increasing befores = 0, and decreasing after)","cited_arxiv_id":null,"evidence_quote":"Contains the proof of the folklore theorem $\\pi_1^{\\mathrm{Morse}}(f,\\ast)\\cong\\pi_1(M,\\ast)$, the invariance baseline the continuation morphisms must reproduce."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the grafted trajectory set-up in a monopole setting that the paper's grafted moduli spaces are modeled on."},{"cited_title":"Translated from the 2010 French original by Reinie Erné","cited_arxiv_id":null,"evidence_quote":"Supplies the interpolation-square gluing framework used to prove functoriality via the angle parameters $\\lambda$ and $\\alpha$."},{"cited_title":"We have a unique critical point ats = 0, which happens to be a minimum, and its limits at±∞ are +∞ (in particular, it is decreasing befores = 0, and increasing after)","cited_arxiv_id":null,"evidence_quote":"Provides standard Morse theory background on unstable manifolds and continuation maps used in the construction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the relative fundamental group $\\pi_1(X,A,\\ast)$ used in Theorem 4.1."}],"review_version":1}