{"id":"7f2f1481-1514-43f6-9ce8-a4355984e4f3","arxiv_id":"2411.18032","paper_version":1,"verdict":"UNVERDICTED","confidence":"HIGH","novelty_score":2.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"An expository survey explaining Milnor invariants from classical links, through welded links, to surface-links and m-dimensional cut-diagrams, with algorithms for computing them.","lead":"This paper is an expository survey of Milnor invariants, a family of link invariants generalizing linking numbers, covering classical links, welded links, surface-links, and higher-dimensional cut-diagrams. It is a translation of a Japanese article written for Sugaku, and is intended to give a broad audience a route into the author's research program.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The invariant ν_C is defined via a Magnus expansion in a quotient group that is not free nilpotent, and the gluing conditions that make cut-diagrams well-defined are deferred to the unpublished [3]; the central claims stand or fall with those omitted details.","rationale":"The paper explicitly identifies itself as an expository translation: 'This is not a research paper, so we shall sometimes sacrifice precision and give rough explanations.' The reader's UNVERDICTED verdict is appropriate because the higher-dimensional claims in Chapters 4 and 5 are quoted from the unpublished preprint [3] and are not proved in the text. My stress-test agrees with that assessment and sharpens one specific technical point: §4.6 defines Milnor invariants via the Magnus expansion of ξ_q(w(l)) in N_q(C,p), but N_q(C,p) is presented with extra commutator relations, so without a specified lift the coefficients are not manifestly well-defined. The text itself routes the missing details to [3]: 'For a detailed definition, see [3, Subsection 1.2.1]' and 'For the detailed definition, see [3]'. This is not an internal inconsistency of the survey, nor an ad hominem concern; it is a verification gap in the paper as a standalone document. If the unpublished preprint supplies the missing gluing conditions and proves well-definedness and the stated invariance theorems, then Corollary 5.3 would be valid. The proposed test checks exactly the well-definedness of ν_C on a simple genus-one case, which is the minimal setting where the non-free quotient relations can matter. Since the reader already flagged the dependence on [3] as the weakest assumption, the verdict should remain UNVERDICTED; no change is needed.","tokens_in":23354,"tokens_out":8411,"duration_ms":79827,"concrete_test":"Obtain preprint [3] and reproduce the definition of ν_C in its §1.2 for the cut-diagram of an unknotted genus-one surface with a single double curve. Compute µ(C,p)(I;l) for the two longitude generators l_1,l_2 using (a) the presentation of Theorem 4.6(1) with two different lifts of ξ_q(w(l)) to the free group F, and (b) the Chen map η_q(w(l)) with two different choices of curves γ_ij. If the coefficient, or the gcd m(C,p)(I_i), changes under either choice, then ν_C is not well-defined as stated and Theorems 4.4–4.5 and Corollary 5.3 are unsupported; if all choices agree, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central construction is not self-contained. §4.4.2 defines 2-dimensional cut-diagrams as triples (Σ,D,f) where f must make 'local gluing can be performed' around branch points and triple points; the actual condition is deferred to [3, Subsection 1.2.1]. §5.1 defines m-dimensional cut-diagrams via a map f satisfying 'certain gluing conditions', again deferred to [3]. The invariant in §4.6 is introduced as a Magnus expansion of ξ_q(w(l)) in N_q(C,p), but Theorem 4.6(1) presents N_q(C,p) as a quotient of the free group F by the relations [α_i, η_q(w(l_ij))] together with Γ_qF. This is not a free nilpotent group, so coefficients of an element ξ_q(w(l)) are not canonical unless a specific lift is chosen. §4.7 proposes to use the Chen map η_q(w(l)) instead, but η_q depends on choices of curves γ_ij, and the stated independence of all choices (Theorem 4.4) is asserted without proof. Since Theorem 4.5 (Roseman invariance) and Corollary 5.3 (concordance invariance of ν_L for m-dimensional links) rest on these definitions, the load-bearing part of the paper is a black box. This is a verification gap, not an accusation: if [3]'s definitions are correct, the concern dissolves.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper is an English translation of an expository article on Milnor invariants, surveying the classical theory, the welded-link extension, and a proposed further extension to surface-links and m-dimensional links via cut-diagrams. The first three chapters present standard material on Milnor invariants for classical and welded links, including Habegger-Lin automorphisms, Chen-Milnor maps, and characterization results in terms of Wk-concordance and self Wk-concordance. The fourth and fifth chapters announce new definitions of Milnor invariants for 2-dimensional and m-dimensional cut-diagrams, claiming Roseman-move invariance (Theorem 4.5) and link-concordance invariance (Corollary 5.3). These central new results are stated without proof and are explicitly deferred to the unpublished preprint [3] by Audoux, Meilhan, and Yasuhara. The paper is openly expository and states that precision is sometimes sacrificed.","tokens_in":23614,"tokens_out":4758,"duration_ms":40754,"significance":"If the announced results are correct, they provide a genuine higher-dimensional analogue of classical Milnor invariants, which would be a valuable contribution to knot theory. The exposition of the classical and welded parts is sound and clearly organized, and the algorithmic treatment via the Chen-Milnor map is a useful feature. However, the significance of the new material cannot be assessed from the manuscript alone: the definitions of 2-dimensional and m-dimensional cut-diagrams rely on gluing conditions deferred to an unpublished source, and the theorems establishing invariance are quoted rather than proved. The paper is therefore best viewed as an expository announcement of results that should appear in a peer-reviewed research article before being relied upon. The manuscript honestly discloses its expository nature and the source of the new results, which is creditworthy, but that does not remove the verification gap.","major_comments":[{"comment":"The definition of a 2-dimensional cut-diagram as a triple (Σ, D, f) requires that f be defined so that 'local gluing can be performed' around each branch point and each triple point, but the actual condition is deferred to [3, Subsection 1.2.1]. All subsequent constructions, including the group G(C, p), the peripheral systems, and the invariants ν_C, depend on this condition. Without the precise gluing condition, the well-definedness of the central objects is not established within the manuscript. Please either include the gluing condition or explicitly state that the paper assumes the results of [3] and adjust the claims accordingly.","section":"Section 4.4.2 and §4.5"},{"comment":"The invariant ν_C is defined through the Magnus expansion of ξ_q(w(l)) in N_q(C, p), which by Theorem 4.6(1) is the quotient of the free group F by the relations [α_i, η_q(w(l_ij))] together with Γ_qF. This group is not a free nilpotent group, so the coefficients of the Magnus expansion of an element are not canonical unless a specific lift is chosen. The manuscript does not specify how the lift is chosen or prove that the coefficients are independent of that choice. This is a load-bearing well-definedness issue for the central definition, and Theorem 4.4, which asserts independence of all choices, is stated without proof.","section":"Section 4.6, Theorem 4.6(1)"},{"comment":"The Chen map η_q(C, p) depends on the choices of the curves γ_ij, as the paper itself notes. The claim that 'Lemma 2.5 (1) holds for η_q(C, p) as well' and Theorem 4.6(2), which identifies ξ_q(w(l)) with η_q(w(l)) modulo the normal closure W, are not proved in the text. Since the computational algorithm and the invariance of ν_C rely on these statements, the manuscript needs either proofs or a precise reference to a publicly available source. As written, the computational method and the invariants it produces are not verifiable from the manuscript.","section":"Section 4.7"},{"comment":"The definition of m-dimensional cut-diagrams and cut-concordance uses a map f satisfying 'certain gluing conditions' deferred to [3], and Theorem 5.2 and Corollary 5.3 are quoted from the unpublished preprint [3]. Because the concordance invariance of ν_L for m-dimensional links is one of the paper's headline claims, the manuscript should either reproduce the relevant definitions and proofs or clearly label these as results proved elsewhere and provide a verifiable reference. As written, the central claims of Chapters 4 and 5 are a black box, and the reader cannot check the arguments.","section":"Sections 5.1–5.2"}],"minor_comments":[{"comment":"There is a typo in the opening sentence: 'In this sction' should be 'In this section'.","section":"Section 4.6"},{"comment":"There is a typo in the sentence introducing the definition for m = 1: 'Threfore' should be 'Therefore'.","section":"Section 5.2"},{"comment":"The notation ρRk(λk) uses the index k both for the component and for the level of the reduced group, which is confusing in statements such as Theorem 3.9. Consider using a different letter for the level, for example ρ_R^r(λ_i).","section":"Section 3.5.1"},{"comment":"In the definition of the group G(C, p), the relation zxz^{-1}y^{-1} is described verbally with reference to Figure 25, but the figure does not clarify which of the arcs x, y, z are labels and which are generators. Please make the convention explicit.","section":"Section 4.5.1"}],"recommendation":"major_revision","confidential_remarks":"This manuscript is an English translation of an expository article originally intended for Sugaku. The new results on surface-links and m-dimensional links are drawn from the author's own unpublished preprint [3]. Before publication in a research-oriented venue, the editor should verify whether [3] is publicly available or under submission; if not, the verification gap is severe. The paper might be more appropriate for an expository journal once the research results are published, or the author should add an appendix containing the omitted definitions and proofs."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [Colleague],\n\nThe short version: this is an English translation of a Sugaku expository article, and it reads like one. If you need a lucid map of Milnor invariants from classical links to welded links to surface-links, this is a good entry point. If you are looking for a research paper with new proofs, this is not it. The last two chapters advertise a genuinely new construction—Milnor invariants for surface-links and m-dimensional links via cut-diagrams—but all the load-bearing details are parked in the unpublished preprint [3] by Audoux, Meilhan, and Yasuhara.\n\nWhat the paper does well: the classical material (Chapters 1–2) is clean and correct, the arrow-calculus and W-tree machinery in Chapter 3 is a nice compendium, and the cut-diagram heuristic in Chapter 4 is intuitive. The author is transparent about the survey nature, explicitly saying it is not a research paper and that precision is sometimes sacrificed.\n\nWhere it is soft: the central new results—Theorems 4.4, 4.5, 5.2, Corollary 5.3—are quoted, not proved. The gluing conditions that make a 2- or m-dimensional cut-diagram well-defined are deferred to [3]. The invariant ν_C is defined via a Magnus expansion in a quotient group that is not free nilpotent, so the coefficients are not obviously canonical; the paper asserts independence (Theorem 4.4) but gives no proof. These are verification gaps, not accusations. If [3] supplies the missing definitions and proofs, the construction likely works; the survey just can't be judged on its own.\n\nOn citations: the self-citations are appropriate—the author owns a large part of this subject, and the deferred results are from his own circle. Not a flaw.\n\nWho is this for: graduate students or researchers who want an overview of Milnor invariants and a peek at the surface-link extension. It is not a paper to referee for a research journal. If it is submitted to an expository venue, a referee should check whether the deferred claims are accurately summarized and whether the cut-diagram definitions are credible from [3]. The real research contribution, [3], is what deserves a full referee report.\n\nRecommendation: if this lands on your desk as a research submission, desk reject—it makes no new claim provable from its own text. If the venue is a survey or proceedings, send it to review with instructions to check the fidelity to [3]. Don't judge the higher-dimensional claims on this paper alone.","headline":"A clear expository survey of Milnor invariants that advertises a new surface-link invariant but leaves all the new proofs in an unpublished preprint—worth reading as a map, not as a research paper.","tokens_in":24242,"tokens_out":3446,"would_cite":false,"duration_ms":30061,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57K10","57K45"],"pacs":[],"model":"deepseek-v4-flash","headline":"Cut-diagrams carry Milnor invariants from classical links to surface-links and higher-dimensional links, yielding new concordance obstructions.","keywords":["Milnor invariants","surface-links","cut-diagrams","Roseman moves","link concordance","welded links","Magnus expansion","higher-dimensional links"],"falsifier":"Compute the Chen-map invariant for the cut-diagram of a specific surface-link such as the spun trefoil, apply a single Roseman move to its diagram, and recompute: any change in $\\nu$ would refute Theorem 4.5. Alternatively, find two m-dimensional links that are known to be link-concordant but whose cut-diagrams yield different Milnor invariants, which would contradict Corollary 5.3.","tokens_in":23054,"feed_emoji":"🪢","tokens_out":8245,"duration_ms":71108,"temperature":0.7,"pith_summary":"This expository paper argues that Milnor invariants, classically defined for collections of circles in 3-space, can be extended to surface-links, meaning closed surfaces smoothly embedded in 4-space, and more generally to m-dimensional links embedded in (m+2)-space. The bridge is the cut-diagram, a kind of net of a link diagram obtained by cutting along under-intersections, which records the gluing data needed to recover the peripheral group. The paper defines Milnor invariants for cut-diagrams, transfers them to surface-links, and states that they are invariant under Roseman moves and under link concordance for every m at least 2. If correct, this gives a higher-dimensional analogue of classical Milnor invariants and a new concordance obstruction for knotted surfaces and beyond. The exposition also covers welded links, where Milnor invariants are completely characterized by the equivalence relations called Wk-concordance and self Wk-concordance.","feed_headline":"Milnor invariants reach surfaces and higher dimensions","feed_subtitle":"Cut-diagram construction turns classical Milnor invariants into concordance obstructions for surfaces and higher links.","key_machinery":"The central object is the cut-diagram, a net of a link diagram obtained by cutting along under-intersections. For surfaces it is a triple (Σ, D, f) consisting of closed surfaces Σ, a 1-dimensional diagram D on Σ whose arcs may end at branch-point markings, and a map f sending each arc of D to a region of Σ − D; the map f encodes which region lies on the other side of each under-intersection. From this data the paper defines a group G(C,p) with relations $zxz^{-1}y^{-1}$ at each arc, a peripheral system with meridian $a_{i0}$ and longitude set $\\Lambda_i = \\{w(l)\\}$ for loops $l$ representing elements of $\\pi_1(\\Sigma_i, p_i)$, and then Milnor invariants by taking coefficients in the Magnus expansion of $w(l)$ and reducing modulo gcds of shorter coefficients. The Chen map $\\eta_q$, defined inductively from the labels met along chosen curves, makes the invariants computable in principle: Theorem 4.6 identifies the nilpotent quotient $N_q(C,p)$ and shows that $\\xi_q(w(l)) = \\eta_q(w(l))$ modulo the relevant normal closure.","core_discovery":"The central claim is that Milnor invariants are not a 3-dimensional phenomenon. For a surface-link L in 4-space, one chooses a diagram and cuts it along its under-intersections to obtain a cut-diagram C = (Σ, D, f); the paper builds a group G(C,p) from the regions of Σ − D and the labels on the arcs of D, equips it with a peripheral system whose ith longitude set Λ_i consists of words w(l) associated to generators of π1(Σ_i, p_i), and reduces Magnus-expansion coefficients modulo gcds of shorter coefficients to obtain integers ν_C(I). These integers are shown to be independent of the chosen base point and of the chosen diagram, so they define invariants ν_L(I) of the surface-link; Theorem 4.5 states that ν_L(I) is unchanged by Roseman moves, the surface analogue of Reidemeister moves. The same construction works for m-dimensional cut-diagrams, and Corollary 5.3 states that ν_L(I) is a link-concordance invariant for every m ≥ 2. For m = 1 the construction recovers the classical and welded Milnor invariants, so the higher-dimensional invariants are presented as a continuation of the classical story rather than a separate theory.","pith_inferences":["The paper leaves implicit that $\\nu_L(I)$ is a probe of the fundamental group of the complement: it notes that the invariants vanish when the fundamental group is trivial, so the first interesting computations would come from knotted surfaces with nontrivial $\\pi_1$, such as spun knots.","The welded-link machinery of Chapter 3, with W-trees and ascending presentations, suggests a testable higher-dimensional analogue: a characterization of when two surface-links have the same Milnor invariants in terms of some surface version of Wk-concordance.","A natural extension would define the same invariants for surface-links with boundary, mimicking string links, which would give a group structure and potentially a Habegger-Lin type classification in dimension two.","If Theorem 4.5 is sensitive enough, $\\nu$ could obstruct unknottedness of surface-links in ways classical invariants cannot, because Roseman moves include birth-and-death and saddle-type changes that have no classical Reidemeister analogue."],"forward_implications":["Milnor invariants $\\nu_L(I)$ are invariants of surface-links: two surface-link diagrams related by Roseman moves have the same values (Theorem 4.5).","For every $m \\geq 2$, $\\nu_L(I)$ is a link-concordance invariant of m-dimensional links, so concordant links share all Milnor invariants (Corollary 5.3).","For $m = 1$ the cut-diagram definition reproduces the classical and welded Milnor invariants, so the higher-dimensional theory reduces to the known theory in the classical case.","The Chen map $\\eta_q$ gives a concrete algorithm for computing $\\nu_C(I)$ of cut-diagrams, extending the classical Magnus-expansion computation to surfaces.","Cut-concordance of cut-diagrams implies equality of $\\nu$ for $m \\geq 2$, so cut-diagrams form a combinatorial model in which concordance questions about higher-dimensional links can be studied (Proposition 5.1 and Theorem 5.2)."],"supporting_citations":[{"why":"Supplies the detailed definition of m-dimensional cut-diagrams, the gluing conditions, and the proofs of Theorems 4.4-4.6, 5.2, and Corollary 5.3 that this article summarizes.","marker":"[3]"},{"why":"Roseman's theorem, stating that two surface-links are equivalent exactly when their diagrams are related by Roseman moves; this is what Theorem 4.5 needs to turn diagram-level invariance into a surface-link invariant.","marker":"[38]"},{"why":"The cited monograph on knotted surfaces and their diagrams, including the lower decker set that the label map f of a cut-diagram generalizes.","marker":"[9]"},{"why":"The original paper defining the invariants via the Magnus expansion, the method that the cut-diagram construction reuses.","marker":"[32]"},{"why":"The paper introducing the link group and the nilpotent-quotient viewpoint on which the peripheral-system definition of Milnor invariants rests.","marker":"[31]"},{"why":"Arrow calculus for welded and classical links, whose W-trees, arrow presentations, and ascending presentations underlie the cut-diagram interpretation and the characterizations of Chapter 3.","marker":"[29]"}],"fun_headline_variants":["Milnor invariants go beyond 3D to surfaces","Surface-links get Milnor invariants via cut-diagrams","Higher-dimensional Milnor invariants from cut-diagrams","Milnor invariants now work for surface-links","Cut-diagrams extend Milnor invariants to higher links"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything in the higher-dimensional chapters rests on the definitions and proofs deferred to an unpublished companion preprint: if the gluing conditions that make an m-dimensional cut-diagram well-defined, or the proofs that $\\nu_C(I)$ is invariant under Roseman moves and concordance, contain a flaw, the central claims of Chapters 4 and 5 collapse.","fun_headline_variants_meta":{"raw":{"variants":["Milnor invariants go beyond 3D to surfaces","Surface-links get Milnor invariants via cut-diagrams","Higher-dimensional Milnor invariants from cut-diagrams","Milnor invariants now work for surface-links","Cut-diagrams extend Milnor invariants to higher links"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000433,"raw_usage":{"total_tokens":2144,"prompt_tokens":816,"completion_tokens":1328,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":432,"completion_tokens_details":{"reasoning_tokens":1248}},"tokens_in":432,"tokens_out":1328,"duration_ms":7394,"temperature":1.0,"reasoning_tokens":1248,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:35:31.793359+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the Chen-map invariant for the cut-diagram of a specific surface-link such as the spun trefoil, apply a single Roseman move to its diagram, and recompute: any change in $\\nu$ would refute Theorem 4.5. Alternatively, find two m-dimensional links that are known to be link-concordant but whose cut-diagrams yield different Milnor invariants, which would contradict Corollary 5.3.","supporting_citations":[{"cited_title":"Audoux, J.B","cited_arxiv_id":null,"evidence_quote":"Supplies the detailed definition of m-dimensional cut-diagrams, the gluing conditions, and the proofs of Theorems 4.4-4.6, 5.2, and Corollary 5.3 that this article summarizes."},{"cited_title":"Roseman, Reidemeister-type moves for surfaces in four-dimensional space , Knot theory, Proceed- ings of the mini-semester, Warsaw, Poland, 1995, (1998), 347–380","cited_arxiv_id":null,"evidence_quote":"Roseman's theorem, stating that two surface-links are equivalent exactly when their diagrams are related by Roseman moves; this is what Theorem 4.5 needs to turn diagram-level invariance into a surface-link invariant."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The cited monograph on knotted surfaces and their diagrams, including the lower decker set that the label map f of a cut-diagram generalizes."},{"cited_title":"Milnor, Isotopy of links , Algebraic geometry and topology, A symposium in honor of S","cited_arxiv_id":null,"evidence_quote":"The original paper defining the invariants via the Magnus expansion, the method that the cut-diagram construction reuses."},{"cited_title":"Milnor, Link groups, Ann","cited_arxiv_id":null,"evidence_quote":"The paper introducing the link group and the nilpotent-quotient viewpoint on which the peripheral-system definition of Milnor invariants rests."},{"cited_title":"Meilhan and A","cited_arxiv_id":null,"evidence_quote":"Arrow calculus for welded and classical links, whose W-trees, arrow presentations, and ascending presentations underlie the cut-diagram interpretation and the characterizations of Chapter 3."}],"review_version":1}