{"id":"e6b622ae-affd-4190-a16b-b3bcc6f4862b","arxiv_id":"2504.20265","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For d≥6 and every k≥1, a new rational decomposition of spaces of homeomorphisms yields R^d-bundles over S^{4k} with nonzero kth rational Pontryagin class, plus a rational section of the stabilization map.","lead":"Building on homotopy-theoretic methods, this paper constructs a rational decomposition of classifying spaces of homeomorphism groups of manifolds. It uses this decomposition to prove that rational Pontryagin classes are nontrivial on R^d-bundles over spheres for all d at least 6, and to produce new tools for detecting characteristic classes of topological bundles.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem A rests on imported [KK24c] embedding-calculus estimates (Lemmas 4.7–4.8); if the cited connectivity bound is off, the pullback decomposition and with it Theorems A, D, E collapse.","rationale":"The paper is carefully written and the proof of Theorem A is surprisingly self-contained once Theorem F is granted: the reduction to the case d = 6, k ≥ 3, the use of Lemma 6.9, the action trick via Proposition 6.4, and the final detection via [GRW17] all check out. The authors' Remark 5.2 explicitly acknowledges the missing fibrewise nilpotency statement for BEmb^θ_∂(M;ℓ_v)^×, but the argument in Remark 5.2(ii) that this nilpotency is unnecessary is plausible: Lemma 2.1 only requires the bottom row of the relevant rationalised pullback square to be 1-connected and nilpotent, and the proof of Proposition 5.5 verifies this for the bottom map after taking fibres, without needing nilpotency of the top-left corner. The family-signature subtlety in Remark 6.7 is also addressed via the canonical summand argument. The genuinely load-bearing point is the reliance on the connectivity and equivalence statements from the authors' earlier work [KK24c]: Lemmas 4.7 and 4.8 are the only bridge from the technically intricate embedding-calculus machinery to the pullback decomposition, and they are cited rather than proved. The reader identified this as the weakest assumption; I agree with that identification, but I do not see an internal error in how the paper uses those results. Hence the verdict remains CONDITIONAL: the main theorems should be accepted only after the cited estimates from [KK24c] are independently verified, especially the exact (k−d+3) connectivity bound in Lemma 4.7(ii) and the pullback square of [KK24c, Section 5.7] used in Lemma 4.8.","tokens_in":74635,"tokens_out":24271,"duration_ms":234338,"concrete_test":"Read [KK24c, Theorem 6.3 and Remark 6.11] and independently re-derive the connectivity estimate used in Lemma 4.7(ii) for the specific case of the bordisms W = ∂_v M × I and W = M with M = W_{g,1}, d = 6, ∂_h M = D^5, and ∂_{vh} M = S^4. Verify that the map ∂_h_{≤k} W → ∂_h W is (k−d+3)-connected for all k, and in particular that at k = ∞ it is an equivalence under the tangential 2-type assumption. If the connectivity is (k−d+2) instead, then the unlabeled horizontal maps in (23) are not equivalences at ∞ and the pullback decomposition Theorem 5.1 fails for d = 6; this would invalidate Theorem A.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central pullback decomposition (Theorem 5.1/Theorem F) is only as secure as the imported embedding-calculus estimates of [KK24c]. In the proof of Theorem 4.1, the dashed filler into T∞Emb^p_∂(M)^× requires Lemma 4.7(iii) (the map ∂_h M → ∂_h_{≤∞} M is an equivalence under tangential 2-type hypotheses) and Lemma 4.8 (the maps ∂_h_{≤k} W → ∂_h_{≤k,p} W and ≤_k W → ≤_{k,p} W are equivalences). Lemma 4.7(ii) cites [KK24c, Theorem 6.3 and Remark 6.11] for a (k−d+3)-connectivity bound, and Lemma 4.8 cites [KK24c, Section 5.7] for a pullback square identifying topological and particle embedding calculus. Neither is reproved here. If the connectivity bound is off by one, or the pullback square in [KK24c, §5.7] fails at a subtle basepoint or component issue, then the square (23) has a non-equivalence on the unlabeled horizontal maps, so the filler in Theorem 4.1—and hence Theorem 5.1, Theorem F, and Theorem A—does not follow. The paper explicitly acknowledges in Remark 4.2(ii) that the T∞-part relies on convergence of topological embedding calculus [KK24c, Theorem 6.4]. This is an external dependency, not an internal contradiction, but it is the single most load-bearing assumption for the paper's central claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs, for smoothable compact d-manifolds with 2-connected interior and boundary, a rational homotopy pullback decomposition for classifying spaces of homeomorphisms and self-embeddings (Theorem F, refined in Theorem 5.1). It applies this decomposition to prove that for d≥6 and k≥1 the kth rational Pontryagin class p_k is nonzero on a topological R^d-bundle over S^{4k} (Theorem A / Theorem 6.8), that the stabilisation map BTop(d)→BTop has a rational section after looping (Corollary B), that BDiff_∂(D^d) has infinitely many nontrivial rational homotopy groups (Corollary C), and that a rational Burghelea–Lashof splitting holds without Postnikov truncation under mild hypotheses (Theorem D/Theorem 6.10). A further application detects certain products of tautological Pontryagin classes (Theorem E). The proof combines a nullhomotopy result for the twice-iterated stabilisation map on SAut(E_{d,Q}) (Theorem G), embedding-calculus estimates imported from the authors' prior work [KK24c], and a new boundary-recovery statement in embedding calculus (Theorem 4.1).","tokens_in":74806,"tokens_out":8216,"duration_ms":81532,"significance":"If the central theorem is correct, the paper solves a problem left open by Weiss and strengthens the Galatius–Randal-Williams stable result in the homotopical direction: it produces sphere bundles with nonzero rational Pontryagin classes in all degrees for d≥6, without proving (and not needing) the corresponding homology surjectivity. The method is novel: the pullback decomposition gives a mechanism to modify vertical tangent bundles of bundles while keeping the underlying fibration fixed, which is the engine behind Theorems A, D, and E. The paper is unusually explicit about its technical assumptions and limitations: Remarks 5.2 and 6.7 flag the fibrewise-nilpotency issue, Remark 4.2(ii) flags the dependence on convergence of topological embedding calculus, and the appendices contain substantial independent material on tensor products of truncated and rational pro-operads. The proof of Theorem G is independent of the Fresse–Willwacher graph-complex route, which is a genuine contribution. The main reservation is that several load-bearing connectivity and pullback statements are cited from the authors' own [KK24c] rather than proved or precisely restated here.","major_comments":[{"comment":"The pullback decomposition in Theorem 5.1 (and hence Theorem F and its applications A, D, E) rests on imported embedding-calculus estimates from [KK24c] that are not reproved in this paper. Specifically, Lemma 4.7(ii) cites [KK24c, Theorem 6.3 and Remark 6.11] for the (k−d+3)-connectivity of ∂^h W → ∂^h_{≤k} W; Lemma 4.8 cites [KK24c, Section 5.7] for the pullback square relating topological and particle embedding calculus; and Lemma 5.3 again uses [KK24c, Theorem 6.3 and Remark 6.11]. The dashed filler in Theorem 4.1, and with it the pullback square (34) in Proposition 5.5, would fail if the cited connectivity bound were off by one or if the pullback in [KK24c, §5.7] had a component/basepoint subtlety. Since these statements are the single most load-bearing external input, the authors should either prove them (perhaps in an appendix) or restate them as explicit, precisely formulated hypotheses of Theorems 5.1 and 6.8, so that the reader can verify the dependency.","section":"§4.3–§4.4, §5.1–§5.2 (Lemmas 4.7, 4.8, 5.3)"},{"comment":"The paper acknowledges in Remark 5.2 that fibrewise nilpotency of BEmb^θ_{∂v}(M;ℓ_v)^× is not established, and argues it is unnecessary. However, the proof of Proposition 5.5 applies Lemma 2.1 to the square obtained by taking fibres to basepoints, and the reader must infer that the bottom row of that fibre square is the map on diagonal fibres of (35) and that it is 1-connected and nilpotent. This is plausible, but the argument is currently compressed into one sentence. Please expand this step to make explicit why Lemma 2.1 applies without any nilpotency assumption on the BEmb-term, and reconcile this with the statement in Remark 5.2(i) that the missing nilpotency can be proved by future work.","section":"§5.2, Remark 5.2, Proposition 5.5"}],"minor_comments":[{"comment":"In the proof of Proposition 3.8, the case SAut≤2(E_{d,Q}) invokes 'the final part of Lemma 3.7(v)', but Lemma 3.7(v) carries an exception for k=2 and degrees i=j=d−1. The exception appears not to contribute to the Bousfield–Kan spectral sequence for π_i of the mapping space because the relevant entry is H^0(S^{d-1}_Q; π_{d-1}). Please add a sentence confirming that the exceptional degree cannot occur, to remove any ambiguity.","section":"§3.2, Proposition 3.8"},{"comment":"The reduction to d=6 and k≥3 is stated correctly, but for k=1,2 the text says p_k evaluates nontrivially on π_{4k}(BSO(6)) without giving a reference. A short citation or one-line argument would improve readability.","section":"§6.3, Proof of Theorem 6.8"},{"comment":"The dimension hypothesis in Theorem D is d≥13, while Theorem 6.10 states a sharper result with d≥8 and additional conditions. It would help the reader to see this relationship stated explicitly in the introduction, since the current phrasing makes Theorem D appear to be a direct corollary rather than a special case of a sharper theorem.","section":"§1, Theorem D / §6.4, Theorem 6.10"}],"recommendation":"major_revision","confidential_remarks":"The central technical inputs are imported from the authors' own prior work [KK24c]: the connectivity estimates of topological embedding calculus (used in Lemmas 4.7 and 5.3) and the topological-vs-particle pullback (used in Lemma 4.8). I could not verify these statements from the present manuscript alone. If [KK24c] is not yet accepted or publicly available, the editor may wish to require that its relevant theorems be stated precisely or that the proofs be included in a companion note. This is a genuine conditionality, not a cosmetic one: an error in [KK24c, Theorem 6.3] would invalidate Theorems A, D, and E as stated. The paper's own innovation and the honesty of its Remark 5.2 are strengths, but the external dependency should be resolved before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The main thing you should know: this paper is the real deal. It closes Weiss's question on rational Pontryagin classes of Euclidean R^d-bundles for d>=6, proving that the classes are nontrivial on sphere bundles (Theorem A), and it gives a rational section of the stabilisation map after looping (Corollary B). The new machinery is the rational pullback decomposition (Theorem F), the nullhomotopy theorem for the double stabilisation of SAut(E_{d,Q}) (Theorem G), and the applications to tautological classes and a Burghelea–Lashof splitting. None of this is a routine repackaging; the tensor-product material on truncated operads in Appendix A and the boundary-recovery results in Section 4 are genuinely new and likely reusable.\n\nThe paper is also honest about its moving parts, which I appreciated. Remark 5.2 flags that fibrewise nilpotency of BEmb^θ_∂(M;·)^× is not established, and the authors explain why the applications don't need it. They also point precisely to which connectivity estimates are imported from their earlier work [KK24c]. That level of transparency is rare.\n\nWhere are the soft spots? The central pullback decomposition in Theorem 5.1 is only as solid as the embedding-calculus lemmas imported from [KK24c]—Lemmas 4.7 and 4.8 cite specific theorems for connectivity and for the topological-vs-particle embedding calculus pullback. Those estimates are load-bearing and not reproved here. I don't see an actual error; the citations are precise and by the same authors, but a referee without deep familiarity with [KK24c] will have to take substantial machinery on faith. The fibrewise nilpotency gap is more of a paperkeeping issue: the authors argue it's unnecessary via the Q-cohomology summand property (Remark 6.7), and that reasoning looks sound, although I wanted it spelled out more fully. The heavy self-citation is legitimate here—they are citing their own prior theorems, not hiding assumptions.\n\nWho is this for? Anyone working on classification of high-dimensional manifolds, characteristic classes of topological bundles, or diffeomorphism groups. The applications section is well written and makes the payoff clear.\n\nRecommendation: send it to referees. The central claim is new and significant, the proofs are detailed, and the main risk—the imported estimates—is exactly what careful refereeing should check. Conditional acceptance, but this deserves serious referee time.","headline":"A substantial, honest paper that resolves Weiss's sphere-detection question for Euclidean bundles in dimensions d>=6; the main theorem sits on imported embedding-calculus estimates, so it deserves refereeing rather than desk rejection.","tokens_in":75521,"tokens_out":1811,"would_cite":true,"duration_ms":21779,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["55R40","57R20","57S05","55P62"],"pacs":[],"model":"deepseek-v4-flash","headline":"For every d≥6 and k≥1, a Euclidean R^d-bundle over the 4k-sphere realizes the k-th rational Pontryagin class.","keywords":["topological Pontryagin classes","Euclidean bundles","homeomorphism groups","rational homotopy pullback decomposition","embedding calculus","stabilisation map","characteristic classes","E_d operad"],"falsifier":"A concrete falsifier would be to compute, for $d=6$ and $k=3$, the evaluation of $p_3$ on the Hurewicz image of $\\pi_{12}(BSTop(6))$ using the paper's pullback square and find it zero, or to exhibit any $d \\ge 6$ and $k \\ge 1$ for which the double-stabilisation map $BSAut(E_{d-2,\\mathbb{Q}}) \\to BSAut(E_{d,\\mathbb{Q}})$ is not nullhomotopic, which would break Theorem G and hence the decomposition.","tokens_in":74259,"feed_emoji":"🧩","tokens_out":16515,"duration_ms":143001,"temperature":0.7,"pith_summary":"The paper establishes that rational Pontryagin classes of Euclidean fibre bundles do not vanish in fixed rank: for all $d \\ge 6$ and $k \\ge 1$, there is a fibre bundle over $S^{4k}$ with fibre $\\mathbb{R}^d$ whose $k$-th rational Pontryagin class is nonzero. Equivalently, the stabilisation map $BTop(d) \\to BTop$ from homeomorphisms of $\\mathbb{R}^d$ to stable homeomorphisms is rationally surjective on homotopy groups after looping, a property that the previously known injectivity on rational cohomology does not by itself supply. The proof works by constructing a rational homotopy pullback decomposition for classifying spaces of homeomorphism groups, which lets one change the vertical tangent bundle of a fibre bundle without changing the underlying fibration. A reader should care because this supplies a new construction method for topological bundles with prescribed characteristic classes and yields infinite families of previously undetected rational homotopy groups of disc diffeomorphism spaces.","feed_headline":"Bundles over spheres realize every rational Pontryagin class for d≥6","feed_subtitle":"It gives a rational splitting of stabilisation after looping, and new rational homotopy groups for disc diffeomorphisms.","key_machinery":"The central object is a rational homotopy pullback square completing the sequence $$BHomeo^\\theta_\\partial(M;\\ell_{1/2\\partial})^{f\\mathbb{Q}} \\to BAut^\\theta_\\partial(TM;\\ell_{1/2\\partial})^{f\\mathbb{Q}} \\to BAut_\\partial(M)_{\\ell_{1/2\\partial}}$$ for smoothable 2-connected compact $d$-manifolds with $d \\ge 5$, 2-connected boundary, and an oriented tangential structure through $BSTop(d-2)$ after rationalisation. The square lets the vertical tangent bundle of a topological $M$-bundle be varied independently of the underlying fibration. Its construction rests on Theorem G: the twice-iterated stabilisation map $BSAut(E_{d-2,\\mathbb{Q}}) \\to BSAut(E_{d,\\mathbb{Q}})$ is nullhomotopic, where $E_{d,\\mathbb{Q}}$ is the rationalised little-$d$-discs operad and $SAut$ is the kernel of the action on the top homology of binary operations; this nullhomotopy makes the rationalised tangential-structure classifying map factor through a contractible piece.","core_discovery":"The paper proves that for every $d \\ge 6$ and $k \\ge 1$, the $k$-th rational Pontryagin class $p_k \\in H^{4k}(BSTop(d); \\mathbb{Q})$ evaluates nontrivially on the Hurewicz image of $\\pi_{4k}(BSTop(d))$, equivalently that some $\\mathbb{R}^d$-bundle over $S^{4k}$ has nonzero $p_k$, and that the stabilisation map $BTop(d) \\to BTop$ admits a rational section after looping. The engine is a rational homotopy pullback decomposition for classifying spaces of homeomorphism groups with tangential structures (Theorem F, proved as Theorem 5.1), obtained by combining topological and particle embedding calculus convergence with a new nullhomotopy of the twice-iterated stabilisation map for automorphisms of the rational $E_d$-operad (Theorem G). From the same decomposition the paper also derives a rational Burghelea–Lashof splitting without Postnikov truncation and a method for detecting tautological classes of topological manifold bundles.","pith_inferences":["If the pullback square is as robust as Theorem F suggests, the same mechanism should construct integral rather than rational Pontryagin classes on sphere bundles whenever an integral version of the nullhomotopy input can be supplied.","Theorem G's nullhomotopy of the double stabilisation of $SAut(E_{d,\\mathbb{Q}})$ may imply a stronger, essentially formal control over the rational homotopy type of automorphism spaces of the $E_d$-operad; the paper deliberately avoids graph-complex methods, leaving this as a natural check.","The $T_\\infty$-boundary discussion points to a new invariant of a manifold's interior recoverable from embedding calculus; testing it on homology spheres with nontrivial fundamental group would map out exactly when boundaries can be recovered from interiors.","The same decomposition could be combined with other tangential structures to detect unstable classes of $BTop(d)$ beyond Pontryagin classes, such as products of classes or classes pulled back from the identity component."],"forward_implications":["For every $d \\ge 6$ there is a rational section, after looping, of the stabilisation map $BTop(d) \\to BTop$, so the latter is surjective on rational homotopy groups.","The disc diffeomorphism space $BDiff_\\partial(D^d)$ for $d \\ge 6$ has infinitely many nontrivial rational homotopy groups, many outside previously computed ranges and in degrees different from configuration-space-integral classes.","A rational Burghelea–Lashof splitting holds without Postnikov truncation for a large class of 2-connected smoothable manifolds, giving topological (not smooth) sections of the forgetful map from fibre-homotopically trivialised fibre bundles to block bundles after looping.","For manifolds containing $S^m \\times S^{d-m}$ as a connected summand, the tautological classes $\\kappa_{p_i p_j}$ are nonzero under explicit degree bounds, and the rational cohomology of the homeomorphism group can grow faster than any polynomial.","The pullback decomposition itself gives a general detection principle: to detect a characteristic class of topological $M$-bundles it suffices to detect the corresponding tautological class on a fibrewise rationalised mapping space."],"supporting_citations":[{"why":"Supplies the topological and particle embedding calculus towers, the convergence and connectivity estimates of Lemma 4.7, and the pullback relating the two calculi of Lemma 4.8, on which Theorem F's decomposition square rests.","marker":"[KK24c]"},{"why":"Introduced the pullback-decomposition strategy for detecting topological Pontryagin classes on sphere bundles that this paper extends to all d≥6 and k≥1.","marker":"[Wei21]"},{"why":"Proved rational injectivity of the stabilisation map BTop(d)→BTop in cohomology, giving the starting fact that p_k is nonzero in H*(BTop(d);Q) for all k.","marker":"[GRW23]"},{"why":"Provides the parametrised Pontryagin–Thom and stable moduli-space input used in the proof of Theorem 6.8 to detect the tautological class κL_k on BDiff^{θ_sm}_∂(W_{g,1}).","marker":"[GRW17]"},{"why":"Supplies the surjectivity theorem for Aut(E_2,Q)→GL(Q) and the pro-rational operad and cyclotomic-action framework used to prove the nullhomotopy of twice-iterated stabilisation (Theorem G).","marker":"[BdBH21]"},{"why":"Computes the coefficient of p_1^k in the Hirzebruch L-class L_k, used to show L_k is nontrivial on τ>3BSO(3) and thus to detect κL_k.","marker":"[BB18]"}],"fun_headline_variants":["Rationally, sphere bundles realize every Pontryagin class","Rationally split stabilisation: topological bundles detect all p_k","Rational pullback for homeomorphism spaces yields new classes","For d>=6, sphere bundles hit every rational Pontryagin class","Homeomorphism groups: rational decomposition reveals sphere-bundle classes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof depends on unproved technical estimates about how well spaces of embeddings of a manifold are approximated by a sequence of simpler spaces; if those estimates fail, the central pullback decomposition and the main theorems collapse.","fun_headline_variants_meta":{"raw":{"variants":["Rationally, sphere bundles realize every Pontryagin class","Rationally split stabilisation: topological bundles detect all p_k","Rational pullback for homeomorphism spaces yields new classes","For d>=6, sphere bundles hit every rational Pontryagin class","Homeomorphism groups: rational decomposition reveals sphere-bundle classes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001544,"raw_usage":{"total_tokens":6166,"prompt_tokens":928,"completion_tokens":5238,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":544,"completion_tokens_details":{"reasoning_tokens":5168}},"tokens_in":544,"tokens_out":5238,"duration_ms":39560,"temperature":1.0,"reasoning_tokens":5168,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:34:24.161395+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete falsifier would be to compute, for $d=6$ and $k=3$, the evaluation of $p_3$ on the Hurewicz image of $\\pi_{12}(BSTop(6))$ using the paper's pullback square and find it zero, or to exhibit any $d \\ge 6$ and $k \\ge 1$ for which the double-stabilisation map $BSAut(E_{d-2,\\mathbb{Q}}) \\to BSAut(E_{d,\\mathbb{Q}})$ is not nullhomotopic, which would break Theorem G and hence the decomposition.","supporting_citations":[],"review_version":1}