{"id":"5db6d949-e9ab-4dc3-a898-32bcbe3a3150","arxiv_id":"2507.23290","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper provides a gap-free proof that subcritical handle attachment leaves wrapped Floer homology invariant, adapting Fauck's cofinal-Hamiltonian fix to the Lagrangian setting.","lead":"This paper gives a corrected, detailed proof that attaching a subcritical handle to a Liouville domain leaves wrapped Floer homology unchanged. It repairs a gap in an earlier proof by adapting a fix originally written for symplectic homology.","discovery_kind":"extension","skeptic_critique":null,"referee_report":{"model":"deepseek-v4-flash","summary":"The paper gives an expository proof of the invariance of wrapped Floer homology under subcritical Weinstein handle attachment (Theorem 1.1). For a Liouville domain M with 2c1(M)=0, exact cylindrical Lagrangians L0,L1 with vanishing Maslov classes, and an isotropic sphere S⊂L1∩∂M with trivial conformal symplectic normal bundle, the theorem asserts HW*(L0,L1;M) ≅ HW*(L0,L1∪_S H^n_k; M∪_S H^{2n}_k). The proof follows Fauck's strategy for symplectic homology: it constructs a cofinal family of Hamiltonians whose extensions to the handle create new Hamiltonian chords only in the Liouville subspace {x=y=0}, proves that their Robin-Salamon/Maslov indices go to infinity with the slope, and then uses a direct-limit comparison of the chain complexes in each fixed degree. Along the way the paper develops the wrapped Floer machinery, Viterbo's transfer morphism, and the Weinstein handle model.","tokens_in":38369,"tokens_out":16606,"duration_ms":183194,"significance":"If the proof is correct, the paper provides a useful and much more explicit treatment of a theorem originally proved by Irie, fixing a known gap concerning the cofinal Hamiltonian family. Its main contributions are the detailed construction of the cofinal family, the generic avoidance lemma, the Morse-Bott index estimate in Section 6.1, and the direct-limit diagram in Section 6.2. The expository value is high, particularly for readers wanting the handle-attachment invariance in a form suitable for applications such as the chord conjecture and the circular restricted three-body problem. However, the theorem itself is not new, and the paper's analytic foundations are partly outsourced to references, so the novelty is limited to the corrected construction and its presentation.","major_comments":[{"comment":"The claim that the interpolated Hamiltonian H satisfies X_H = C_x X_x − C_y X_y + C_z X_z with C_x, C_y, C_z > 0 is not established. Lemma 5.4 only shows that the error term (ψ−φ)β′R is small; it does not show that adding this term to the positive-coefficient combination preserves the signs of the coefficients. In particular, the coefficient of X_y in X_{ψ_δ} is −(1−(1+ε)g′), which can be as small as ε/(1+2ε), and the error term contributes −(ψ−φ)β′ c_y X_y for a positive c_y; without a quantitative bound comparing the error to this lower bound, the Lyapunov computation dL/dt = (C_x x + C_y y)/2 may fail, and the conclusion that all new chords lie in {x=y=0} would be unjustified. Since the index calculation in §6.1 and the direct-limit argument in §6.2 both rely on this conclusion, a verified sign-preserving interpolation (or a different interpolation scheme) is required.","section":"§5.1.2, Eq. (9)"},{"comment":"The proof of Lemma 5.2 does not justify the step from the dimension/transversality statement to the assertion about Reeb chords from Λ0 to Λ1 entering H_δ. The transversality argument shows that a generic contact form has no Reeb chord from Λ0 to S, but a chord from Λ0 to Λ1 that enters a thin handle need not end on S; one must use compactness of chords of period ≤ a to pass from avoidance of Λ0×S to avoidance of a neighborhood of S in Λ1, and the dimension count should be applied to Y×[0,a] rather than to Y×R_+. This is a local fix, but the lemma is a central input to the cofinal family construction and should be proved explicitly.","section":"§5.1, Lemma 5.2"},{"comment":"The case of distinct Lagrangians L0 and L1 is dismissed in Remark 5.1 with the assertion that 'the only 1-chords of H are the chords in M'. This is not immediate: a chord starting on L0 and ending on the new handle H^n_k is not a Reeb chord from Λ0 to Λ1, so Lemma 5.2 as stated does not rule it out, and the Lyapunov argument does not control L along L0 because L need not vanish there. Since Theorem 1.1 is stated for general L0,L1, the distinct case either needs a full proof (e.g., an avoidance statement for Λ0×(H^n_k∩Σ_δ)) or the theorem should be stated only for L0=L1 with the general case delegated to a reference.","section":"§5.1 Remark 5.1 and §6.2"}],"minor_comments":[{"comment":"There are numerous typographical errors, e.g., 'Translted' in reference [AD14], 'OVER VIEW' in [Wen], and 'Arnol‘d'; these should be corrected.","section":"Throughout"},{"comment":"The commutative diagram is garbled; the lower-right corner should presumably be HF_k(L0,L1∪H^n_k,H_{i+1}), not HW_k(L0,L1∪H^n_k,H_i).","section":"§6.2"},{"comment":"The domain of F_α is written as M×R_+, but the dimension count (2n) indicates Y×R_+ is intended; this should be clarified.","section":"§5.1, proof of Lemma 5.2"},{"comment":"The symbol δ is used both for the handle parameter and for the error tolerance; this is confusing and should be renamed.","section":"Lemma 5.4"},{"comment":"The sentence 'we have already shown just shown that w:C→M is non-constant' contains a duplicated phrase.","section":"§3.3.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is an expository contribution fixing a known gap; the main theorem is due to Irie. The central new construction—the cofinal family with controlled index growth—is valuable, but the positivity claim in Eq. (9) and the treatment of the distinct-Lagrangian case are load-bearing and need repair. If those are fixed, the paper would be a useful reference. The journal should also consider whether the expository nature and the reliance on prior work fit its scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a careful, honest write-up of Irie's theorem with a real repair of the cofinal-family gap. I think the proof is very likely correct, but there is one spot in Section 5.1.2 where the paper asserts something it has not checked.\n\nWhat is new is not the theorem; it is the adaptation of Fauck's symplectic-homology fix to wrapped Floer homology. The cofinal family construction in Section 5.1, the generic avoidance Lemma 5.2, the Morse-Bott index calculation in Section 6.1, and the direct limit diagram in Section 6.2 are laid out in real detail. The index control of the newly created chords is the right idea, and the computation is transparent. The paper also gives a genuinely useful self-contained introduction to wrapped Floer homology, Maslov grading, and Viterbo's transfer morphism. For an expository article, that is a real service.\n\nThe main mathematical soft spot is equation (9). Lemma 5.4 bounds the interpolation error term (psi-phi) beta' R; it does not show that this error preserves the positive-coefficient structure C_x X_x - C_y X_y + C_z X_z. The paper moves from 'X_H is close to the interpolation' to 'X_H satisfies (9)' as if positivity of C_x, C_y, C_z were automatically stable. Since the error term lies along the Reeb direction, which in this model is itself a positive combination of X_x, X_y, X_z, the issue reduces to sign control of psi-phi, and the paper does not provide that control. This is not necessarily fatal: a sharper choice of beta or a smaller delta might well fix it. But it is exactly the kind of implicit step the paper is supposed to be repairing, and it should be spelled out.\n\nOther soft spots are less serious. Several analytic foundations, including transversality, gluing, and some maximum-principle arguments, are cited rather than proved; for this genre that is acceptable. The Maslov-class hypothesis on the surgered Lagrangian L1 union H^n_k is also not explicitly verified; the grading of the wrapped Floer chain complex on the surgered side is assumed rather than checked. I would not treat either as a load-bearing flaw. No circularity problem appears: the self-references occur in applications and comments, not in the engine of the proof.\n\nThis paper is for people who want a quotable, readable proof of the subcritical handle attachment theorem and who need the corrected cofinal family. It deserves a serious referee. The referee should focus on Section 5.1.2 and ask for the missing check of equation (9); after that, the paper is publishable.","headline":"A genuinely useful repair of a known theorem, mostly expository, with one interpolation step that needs an explicit check before the fix is airtight.","tokens_in":38773,"tokens_out":2852,"would_cite":true,"duration_ms":34179,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["53D40","53D35"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that subcritical handle attachment leaves wrapped Floer homology unchanged, and repairs a gap in the earlier cofinal-Hamiltonian construction.","keywords":["wrapped Floer homology","subcritical handle attachment","Weinstein handle","cofinal Hamiltonian family","Maslov index","Liouville domain","Reeb chords","direct limit"],"falsifier":"Compute the coefficient functions $C_x,C_y,C_z$ for the vector field of the interpolated Hamiltonian $H=\\hat H+(h_{\\Sigma_\\delta}-\\hat H)\\beta(h_{\\Sigma_\\delta}+2)$ using the estimate in Lemma 5.4; if any of these coefficients is not strictly positive on an open set, then the argument that all new chords lie in $x=y=0$ fails and the index control is lost.","tokens_in":38133,"feed_emoji":"📐","tokens_out":16665,"duration_ms":165534,"temperature":0.7,"pith_summary":"Wrapped Floer homology is a Lagrangian analogue of symplectic homology: its chain complex is generated by Hamiltonian chords between two exact cylindrical Lagrangians in a Liouville domain, and its differential counts solutions of Floer's equation. The paper's central claim is the invariance theorem: under the standard index assumptions $2c_1(M)=0$ and vanishing Maslov classes, attaching a subcritical Weinstein $k$-handle $H^{2n}_k$ to the Liouville domain along an isotropic sphere in one Lagrangian, while attaching its imaginary part $H^n_k$ to that Lagrangian, leaves $HW^*(L_0,L_1;M)$ unchanged. The theorem had been proved before, but the earlier proof contained a gap in the choice of a cofinal family of Hamiltonians, and this paper repairs that gap. The repair allows the cofinal Hamiltonians to create additional chords inside the handle, provided their Robbin–Salamon indices grow with the Hamiltonian slope; such chords disappear from any fixed degree of the direct limit. A reader should care because wrapped Floer homology is a computable invariant of Liouville domains and their Lagrangians, and knowing it is unchanged under subcritical surgery makes transfer arguments valid.","feed_headline":"Wrapped Floer homology survives subcritical handle attachment","feed_subtitle":"A corrected proof controls the indices of newly created chords, closing a gap in the original argument.","key_machinery":"The load-bearing object is the cofinal family $H_i=a_iH_{\\delta_i}+2a_i+b_i$ on $M\\cup_S H^{2n}_k$, where $H$ is the Hamiltonian obtained by interpolation between the handle's defining function $\\psi_\\delta$ and the radial model $h_{\\Sigma_\\delta}$. Lemma 5.4 constructs the interpolation so that the vector field keeps the positive-coefficient form $X_H=C_xX_x-C_yX_y+C_zX_z$ with $C_x,C_y,C_z>0$; this is what confines all new chords to the Liouville subspace $x=y=0$ through the Lyapunov function $L=\\sum_{i=1}^k x_i y_i$. The handle itself is the standard Weinstein $k$-handle $H^{2n}_k=\\{\\phi\\ge -1\\}\\cap\\{\\psi_\\delta\\le -1\\}\\subset \\mathbb{C}^k\\times\\mathbb{C}^{n-k}$, and its imaginary part $H^n_k$ is the subset $x_i=0$ that is glued to the Lagrangian $L_1$. The explicit Robbin–Salamon index calculation on $x=y=0$, splitting into $k$ non-rotating and $n-k$ rotating directions, is what shows the new chords have index growing linearly with the slope $a$.","core_discovery":"The central content is a corrected proof of Theorem 1.1. Starting from a cofinal family on $M$ that is linear near the boundary, the paper extends each Hamiltonian across the handle by interpolating between the handle's defining function and a radial model, using an interpolation lemma designed to keep the Hamiltonian vector field in the form $X_H=C_xX_x-C_yX_y+C_zX_z$ with $C_x,C_y,C_z>0$ (equation (9)). This positive-coefficient form, together with the Lyapunov function $L=\\sum_{i=1}^k x_i y_i$, forces every newly created Hamiltonian chord to lie in the Liouville subspace $x=y=0$. There the dynamics is explicit, and the Robbin–Salamon index of a chord is $\\mu_{\\mathrm{RS}}(x)=\\frac{n}{2}+(n-k)(\\frac{aC_z}{2\\pi}-\\frac{1}{2})$, so after resolving degeneracies the resulting Maslov indices tend to $+\\infty$ as the slope $a$ grows. In each fixed degree the chain complexes eventually have the same generators, and passage to the direct limit gives the isomorphism.","pith_inferences":["The same interpolation-and-index-control mechanism should extend to partially wrapped Floer homology and to Weinstein sectors, where subcritical handle attachment is expected to be invariant; the paper treats only exact cylindrical Lagrangians in Liouville domains.","The index formula suggests that in a filtered wrapped Floer complex the handle contributes a cluster of generators whose indices grow linearly with the slope; computing filtered groups for the model handle would test this.","Because the proof runs over $\\mathbb{Z}_2$, carrying the isomorphism to integer coefficients would require orienting the moduli spaces of Floer strips; the index-cluster bound does not by itself settle orientations."],"forward_implications":["Wrapped Floer homology is unchanged by subcritical handle attachment: any Liouville domain pair related by such surgery has isomorphic $HW^*$, so invariants computed before surgery transfer directly.","For a single Lagrangian $L=L_0=L_1$, the relative condition $2c_1(M,L)=0$ is sufficient, and the isomorphism simplifies to $HW_*(L;M)\\cong HW_*(L\\cup_S H^n_k; M\\cup_S H^{2n}_k)$.","The new chords created by the repaired cofinal family have Maslov index tending to $+\\infty$ with the slope, so in any fixed degree they cannot affect the direct limit.","When $L_0$ and $L_1$ are distinct, no new $1$-chords are created by the handle at all, giving the invariance in that case without index control."],"supporting_citations":[{"why":"states the invariance theorem for wrapped Floer homology and contains the cofinal-family construction whose gap this paper repairs.","marker":"[Iri13]"},{"why":"supplies the strategy, adapted here, of allowing extra chords and controlling their Robbin–Salamon index, plus the maximum principle used in Lemma 4.1.","marker":"[Fau20]"},{"why":"the doctoral thesis where the gap in the earlier construction was found and resolved for symplectic homology; the paper follows its analysis.","marker":"[Fau16b]"},{"why":"introduced the effect of handle attachment on symplectic homology and contains one of the flawed cofinal-family claims, its Lemma 2.5, that this paper fixes.","marker":"[Cie02]"},{"why":"defines contact surgery along an isotropic sphere and gives the standard model of the handle $H^{2n}_k$ and its imaginary part $H^n_k$.","marker":"[Wei91]"},{"why":"defines the Robbin–Salamon index whose explicit calculation on the subspace $x=y=0$ controls the indices of the new chords.","marker":"[RS93]"},{"why":"gives the bound that controls how the index shifts when a degenerate chord is perturbed into non-degenerate ones.","marker":"[CFHW96]"}],"fun_headline_variants":["Gap fixed: wrapped Floer homology unchanged by subcritical handles","Subcritical handle attachment: invariance proof corrected","Wrapped Floer homology invariant under subcritical handles, gap closed","Invariance proof gap fixed for wrapped Floer homology"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof's control of newly created chords depends on the interpolated Hamiltonian's vector field keeping the positive-coefficient form $X_H=C_xX_x-C_yX_y+C_zX_z$ with $C_x,C_y,C_z>0$; the paper does not explicitly check that the interpolation error term from Lemma 5.4 preserves this form, and if it did not, the confinement of chords to $x=y=0$ and the index bound would fail.","fun_headline_variants_meta":{"raw":{"variants":["Gap fixed: wrapped Floer homology unchanged by subcritical handles","Subcritical handle attachment: invariance proof corrected","Wrapped Floer homology invariant under subcritical handles, gap closed","Invariance proof gap fixed for wrapped Floer homology"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001657,"raw_usage":{"total_tokens":6533,"prompt_tokens":856,"completion_tokens":5677,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":472,"completion_tokens_details":{"reasoning_tokens":5607}},"tokens_in":472,"tokens_out":5677,"duration_ms":42723,"temperature":1.0,"reasoning_tokens":5607,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T10:52:54.821139+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the coefficient functions $C_x,C_y,C_z$ for the vector field of the interpolated Hamiltonian $H=\\hat H+(h_{\\Sigma_\\delta}-\\hat H)\\beta(h_{\\Sigma_\\delta}+2)$ using the estimate in Lemma 5.4; if any of these coefficients is not strictly positive on an open set, then the argument that all new chords lie in $x=y=0$ fails and the index control is lost.","supporting_citations":[],"review_version":1}