{"id":"d2707a60-c4ca-4800-bcd8-e680a5752739","arxiv_id":"2608.07362","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Hochschild-Kostant-Rosenberg maps twisted by the square root of the Todd class produce an isomorphism of calculi between Tamarkin-Tsygan and Cartan calculi on every dg manifold.","lead":"This paper proves that on any differential graded manifold, the Hochschild-Kostant-Rosenberg maps corrected by the square root of the Todd class give an isomorphism between the Tamarkin-Tsygan calculus and the Cartan calculus. It resolves a conjecture from 1998 and unifies Duflo's Lie algebra theorem with Kontsevich's complex manifold theorem.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The descent from the Fedosov calculus to the dg-manifold calculus relies on the unpublished Theorem 2.5 of [30]; if its contractions fail to preserve the full calculus operations, Theorem A is not established.","rationale":"I read the paper in good faith. The overall architecture is plausible: prove a Duflo-Kontsevich-type isomorphism for the Fedosov dg Lie algebroid, then descend to the original dg manifold using the contractions of [30]. The internal computations and the algebra of L∞ morphisms in the body of the paper are detailed and cohere with the stated results, and the proof of Proposition 3.7 reducing the twisted Taylor coefficients to the standard flat case via injectivity/surjectivity is a sound strategy. My concern is not an internal inconsistency but the decisive black-box dependence: Theorem 2.5 of [30] is asserted without proof here, and the descent step would collapse if the injections preserved only the linear structure and not the full calculus operations. This matches the reader's weakest_assumption. I also note Propositions B.3/B.4 are used without proof, but they are somewhat more standard and backed by published references. A minor secondary ambiguity is that Theorem A states 'any dg manifold' while Proposition 2.7 and the Fedosov setup assume finite-dimensionality; this should be clarified but does not affect my main concern. On balance, the verdict CONDITIONAL is appropriate: the theorem is credible but not fully verifiable without the missing proofs.","tokens_in":70214,"tokens_out":15753,"duration_ms":136180,"concrete_test":"Obtain the preprint [30] and verify the calculus-preservation identities for the injections in Theorem 2.5 in the flat case M = R^n with Q = 0 and the standard flat connection, where the Fedosov algebroid and the maps τ̆^♮, σ̃^♮, h̃^♮ are explicit. Check at the cochain level that τ̆^♮([X,Y]) = [τ̆^♮(X),τ̆^♮(Y)] for the Schouten bracket, τ̆^♮(i_X ω) = i_{τ̆^♮(X)}τ̆^♮(ω), τ̆^♮(L_X ω) = L_{τ̆^♮(X)}τ̆^♮(ω), and the analogous identities for the Hochschild-side operations ⌣, ⟦−,−⟧, i, L, B. If any one of these strict identities fails in this basic case, the descent step is invalid and Theorem A is unproven.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim, Theorem A, asserts an isomorphism of calculi at cohomology level. The proof reduces the dg-manifold statement to the Fedosov-level Proposition 3.3 by composing with the calculus-preserving contractions of Theorem 2.5 of the unpublished preprint [30]. The final step of the proof of Theorem A uses the fact that 'the vertical maps are isomorphisms of calculi (by Proposition 3.3 and Theorem 2.5)'. This requires the injections τ̆^♮ in (34)-(38) to be strict morphisms of calculi: they must preserve the differentials and also the Gerstenhaber bracket, the wedge and cup products, the contraction and Lie derivative actions, and the Connes-Rinehart operator B, at the cochain level. The present paper contains no proof of Theorem 2.5, and [30] is a self-cited unpublished preprint. If any of these compatibility identities fails, the cohomology-level maps could still be vector-space isomorphisms but would not be calculus isomorphisms, so Theorem A would not follow. Secondary but similarly load-bearing, Propositions B.3 and B.4 are imported from the literature without proofs; they determine the key twisted Taylor coefficients hkr∘(Â)^{1/2} and (Â)^{1/2}∘HKR used in Proposition 3.7.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a Duflo--Kontsevich-type theorem for Tamarkin--Tsygan calculi of dg manifolds: the HKR maps twisted by the square root of the Todd class induce an isomorphism of calculi between the Cartan calculus and the Tamarkin--Tsygan calculus at cohomology level. The proof proceeds through formality theorems for Fedosov dg Lie algebroids, with explicit Kontsevich--Shoikhet-type morphisms, and then descends to the original dg manifold using contraction data imported from the companion preprint [30]. The paper also states and proves a formality theorem for dg manifolds, Theorem B, whose first coefficients are the twisted HKR maps.","tokens_in":70476,"tokens_out":6365,"duration_ms":64381,"significance":"If correct, the main result confirms the Kontsevich--Shoikhet conjecture and unifies the Duflo--Kontsevich theorem for Lie algebras with the Kontsevich theorem for complex manifolds, at the level of full calculus structures rather than only Gerstenhaber algebras. The paper is carefully structured, with explicit formulas for the HKR maps, the Todd class via Atiyah classes, and the relevant L-infinity morphisms; the homotopy transfer arguments in Appendix A are detailed and several lemmas are proved in full. The Todd class is defined independently through Atiyah classes, so the main theorem is not circular. However, the proof is substantially conditional on unpublished companion results, and the extent of the finite-dimensionality assumptions needs clarification.","major_comments":[{"comment":"The central descent step relies on Theorem 2.5, imported from the unpublished preprint [30] without proof. In the proof of Theorem A, the statement 'the vertical maps are isomorphisms of calculi (by Proposition 3.3 and Theorem 2.5)' is exactly where the calculus-level compatibility of the injections τ̆^♮ is needed: Theorem 2.5(i) and (iii) assert preservation of ∧, [−,−], i, L, d, ⌣, ⟦−,−⟧, and B at cochain level. Since [30] is a self-cited unpublished preprint and no proof of these compatibilities is given here, the descent from the Fedosov calculus to the dg manifold calculus is not self-contained. If those compatibilities fail, the cohomology-level maps could still be vector-space isomorphisms but not calculus isomorphisms, so Theorem A would not follow. The paper should either include a proof of Theorem 2.5 or explicitly state Theorem A as conditional on the results of [30].","section":"§2.2.4 and §4.2, proof of Theorem A"},{"comment":"Theorem A and the abstract claim the result for 'any dg manifold', but several load-bearing ingredients are stated only for finite-dimensional dg manifolds. Proposition 2.7 is explicitly for a finite-dimensional dg manifold, Section 3 opens with 'Given a finite-dimensional dg manifold', and the Fedosov construction in §2.2.1 uses local charts with finitely many virtual coordinates. If the paper's convention is that all dg manifolds are finite-dimensional, this should be stated where 'dg manifold' is first defined; otherwise the proof does not support the stated generality. This is not merely cosmetic, because the HKR quasi-isomorphisms in Proposition 2.7 and the Fedosov descent in Section 4 are the mechanisms that produce the twisted maps in Theorem A.","section":"Theorem A and §2.4.1, Proposition 2.7"},{"comment":"Proposition 3.7, which identifies the first Taylor coefficient 𝔘₁ and the zeroth coefficient 𝔖₀ with the Â-twisted HKR maps, depends on Equation (78). That equation is asserted to follow from Propositions B.3 and B.4, but those propositions are stated without proof and without a precise derivation of their graded/trivialized versions from the cited sources [50,53]. Since these coefficients are the cochain-level incarnations of the maps that later become the Todd-twisted isomorphisms, the computation is load-bearing. The authors should either prove Propositions B.3 and B.4 or give a detailed derivation indicating exactly how the statements in [50,53] imply the graded versions used here.","section":"§3.2, Proposition 3.7 and Appendix B"}],"minor_comments":[{"comment":"The Koszul sign in the definition of the cohomological HKR map is not spelled out; since signs play a role in later computations, please state the sign convention explicitly or refer to a formula where it is fixed.","section":"§2.4.1, Eq. (41)"},{"comment":"The notation 𝒩 = 𝑇ℳ[1] ⊕ 𝑇̂ℳ for what appears to be a fibred product over ℳ is confusing; using ⊕ normally denotes a direct sum of vector bundles. Please clarify the fibred-product nature of this construction.","section":"§2.2.1"},{"comment":"In the proof of Lemma 4.2, the equalities (td^can)^{-1/2} ∧ L_𝒬ζ = L_𝒬ξ and (L_𝒬ζ) ∧ (td^can)^{-1/2} = L_𝒬η are used without showing the sign conventions; a one-line calculation using the derivation property of L_𝒬 would make the homotopy formulas transparent.","section":"Lemma 4.2"},{"comment":"The paper defines 'graded' but not 'dg manifold' precisely, and finite-dimensionality is first imposed only in Proposition 2.7. Please define the class of dg manifolds under consideration at the beginning and keep it consistent with the statements of Theorems A and B.","section":"Introduction, 'dg manifold' definition"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's main theorem is announced in a form that depends heavily on two unpublished companion papers, [29] and [30], by the same authors. In particular Theorem 2.5 of [30] supplies the calculus-preserving contractions that make the whole descent in Section 4 work. I would encourage the editors to ask the authors to make those results available, or to include their proofs in an appendix, before final acceptance. The overbroad 'any dg manifold' wording also deserves correction, since the proof appears to require finite-dimensionality."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this paper proves the calculus-level Duflo–Kontsevich theorem for dg manifolds, not just the cohomology-level Gerstenhaber statement announced in the authors’ CRAS note. The full isomorphism of Tamarkin–Tsygan calculi, with all operations included, is a genuine advance, and Theorem B is supplied with a detailed proof rather than a sketch. The architecture—Fedosov dg Lie algebroid, leafwise Kontsevich/Shoikhet formality, twisting, then descent—is coherent, and the homotopy-transfer appendix is carefully written. I found no internal contradiction and no sign of fitting, so this is serious work.\n\nThe soft spots are real but not demonstrated flaws. The most important is the descent step: it relies on Theorem 2.5 of [30], an unpublished preprint by the same authors. The reduction to the Fedosov-level statement collapses if the contractions in (34)–(38) do not preserve the full calculus operations at cochain level, and the present paper does not prove that theorem; it only cites it. That is a load-bearing external dependency, not a minor gap. Second, Propositions B.3 and B.4, which fix the twisted Taylor coefficients as hkr∘Â^{1/2} and Â^{1/2}∘HKR, are imported from [50,53] without proof. They are standard to the expert, but since the main theorem turns on those coefficients, a referee should see the argument or at least a precise statement. Third, Theorem A says “any dg manifold,” while Proposition 2.7 and Section 2.3 explicitly assume finite-dimensional dg manifolds. The mismatch is probably a tacit standing assumption, but it should be stated.\n\nWho gets value: deformation quantizers, dg-geometry people, and anyone relying on the Kontsevich–Shoikhet conjecture. It is a serious long paper, not a sketch, and it deserves referee time. My recommendation: send it to peer review. Require the authors to make [30] available or fold the needed parts of Theorem 2.5 and Propositions B.3/B.4 into the paper before acceptance.","headline":"Proves the full calculus-level Duflo–Kontsevich isomorphism for dg manifolds; a real proof, but load-bearing dependencies on unpublished companion work should be resolved in review.","tokens_in":71003,"tokens_out":2494,"would_cite":true,"duration_ms":27361,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["16E40","53D55","58A50"],"pacs":[],"model":"deepseek-v4-flash","headline":"For every dg manifold, the HKR maps twisted by the square root of the Todd class give an isomorphism of calculi between the Cartan and Tamarkin–Tsygan models.","keywords":["dg manifolds","Tamarkin–Tsygan calculus","Cartan calculus","Duflo–Kontsevich isomorphism","Todd class","Hochschild–Kostant–Rosenberg maps","Fedosov dg Lie algebroids","formality theorems"],"falsifier":"Take a dg manifold whose Atiyah cocycle is nonzero, for instance the affine space $\\mathbb{R}^{0|2}$ with a homological vector field $Q$ containing a nonzero quadratic term, and evaluate the zeroth Taylor coefficient $(\\operatorname{td}^{1/2}\\circ\\mathrm{HKR})$ on a one-polyjet $\\zeta$. If the cohomology class of $d_{\\mathrm{DR}}((\\operatorname{td}^{1/2}\\circ\\mathrm{HKR})(\\zeta))$ is not equal to the class of $(\\operatorname{td}^{1/2}\\circ\\mathrm{HKR})(B\\zeta)$, where $B$ is the Connes–Rinehart operator, then the claimed calculus isomorphism fails; the theorem predicts equality for every such $\\zeta$ and every such $Q$.","tokens_in":70028,"feed_emoji":"","tokens_out":11064,"duration_ms":97249,"temperature":0.7,"pith_summary":"This paper proves that the two natural cohomology-level calculi attached to any differential graded (dg) manifold are isomorphic: the Cartan calculus of polyvector fields and differential forms, and the Tamarkin–Tsygan calculus of polydifferential operators and polyjets. The bridge is the Hochschild–Kostant–Rosenberg (HKR) map, corrected by the square root of the Todd class of the dg manifold, acting by contraction on polyvector fields and by wedge multiplication on differential forms. The theorem states that the corrected maps preserve every calculus operation, including cup products, Gerstenhaber brackets, contractions, Lie derivatives, the Connes–Rinehart operator, and the differentials, not merely the underlying chain structures. At cohomology level this confirms the Kontsevich–Shoikhet conjecture stated in 1998. The payoff is unification: the classical Duflo theorem for Lie algebras and the Kontsevich theorem for complex manifolds become two special cases of one dg-manifold statement.","feed_headline":"Todd-class twist makes dg-manifold calculi isomorphic","feed_subtitle":"For every dg manifold, Hochschild and Cartan cohomology carry the same operations, uniting two classical Duflo-type theorems.","key_machinery":"The load-bearing object is the square root of the Todd class $(\\operatorname{td}_{(\\mathcal{M},Q)})^{1/2}$, defined as the square root of the Berezinian of the Atiyah class of the tangent dg Lie algebroid; it is a degree-zero element of the differential-form space and acts on one leg of the calculus by contraction and on the other by multiplication. Around this class the proof builds two formality morphisms: a Kontsevich-type $L_\\infty$ quasi-isomorphism from polyvector fields to polydifferential operators, and a Shoikhet-type quasi-isomorphism of $L_\\infty$ modules from polyjets to differential forms, both twisted by the homological vector field. The passage to dg manifolds is made through Fedosov dg Lie algebroids, whose calculus is connected to the tangent dg Lie algebroid by contractions preserving all operations, and through a divergence identity that converts the $\\hat{A}$ class of the Fedosov algebroid into the Todd class.","core_discovery":"The central discovery is Theorem A: for every dg manifold $(\\mathcal{M},Q)$, the pair of maps $(\\operatorname{hkr}\\circ(\\operatorname{td}_{(\\mathcal{M},Q)})^{1/2},\\ ((\\operatorname{td}_{(\\mathcal{M},Q)})^{1/2}\\circ\\mathrm{HKR})^{-1})$ is an isomorphism of calculi from $\\mathrm{Cal}_C(\\mathcal{M},Q)$ to $\\mathrm{Cal}_H(\\mathcal{M},Q)$, where $(\\operatorname{td}_{(\\mathcal{M},Q)})^{1/2}$ is the square root of the Todd class of the dg manifold. At the cohomology level this gives a full calculus isomorphism, so the Tamarkin–Tsygan and Cartan calculi are not just quasi-isomorphic as complexes; they carry the same cup products, Gerstenhaber brackets, contractions, Lie derivatives, Connes–Rinehart operators, and differentials. The theorem is proved through a formality theorem for dg manifolds (Theorem B) whose first Taylor coefficient is the Todd-twisted HKR map, and it specializes to the Duflo–Kontsevich theorem for Lie algebras and to the Kontsevich theorem for complex manifolds.","pith_inferences":["One extension not pursued in the paper is that the same Todd-class correction should mediate comparisons between other pairs of models for polyvector fields on dg manifolds, such as negative cyclic chains and equivariant forms.","A testable consequence of the proof's construction is that the resulting calculus isomorphism should be independent of the auxiliary affine connection up to explicit homotopies; the affine dg-manifold example in the paper is a concrete setting where those homotopies could be written down.","A natural next test case is the derived intersection dg manifold $E[-1]$ associated to a section of a vector bundle, where the Todd class should reduce to a known characteristic class and the twisted HKR maps can be computed explicitly."],"forward_implications":["For any dg manifold, the Hochschild and Cartan cohomologies agree as full Gerstenhaber algebras with compatible module and differential structures, so computations in either model produce the same invariants.","The theorem subsumes the Duflo–Kontsevich isomorphism for Lie algebras, obtained when the dg manifold is the shift of a Lie algebra with the Chevalley–Eilenberg differential.","It subsumes the Kontsevich theorem for complex manifolds, obtained from the Dolbeault dg manifold of a complex manifold.","Theorem B supplies an $L_\\infty$ formality quasi-isomorphism for every dg manifold, with first Taylor coefficient equal to the Todd-twisted HKR map and a compatible module quasi-isomorphism for the chain side; the authors expect it to be applicable to deformation quantization of $0$-shifted derived Poisson manifolds.","The Todd class of a dg manifold acts as a single characteristic class encoding both the Duflo element of a Lie algebra and the ordinary Todd class of a complex manifold."],"supporting_citations":[{"why":"Formulates the Kontsevich–Shoikhet conjecture, confirmed at cohomology level by Theorem A, that Todd-twisted HKR maps give a Duflo-type isomorphism for dg manifolds.","marker":"[43]"},{"why":"Supplies Theorem 2.5, the calculus-preserving Fedosov contractions between the tangent and Fedosov dg Lie algebroids, which let the proof descend from Fedosov calculi to dg-manifold calculi.","marker":"[30]"},{"why":"Kontsevich's formality theorem for polyvector fields to polydifferential operators on a trivialized graded manifold, used leafwise to build the Fedosov formality morphisms.","marker":"[23]"},{"why":"Shoikhet's formality theorem for Hochschild chains gives the $L_\\infty$ module morphism from polyjets to differential forms whose zeroth Taylor coefficient is the twisted HKR map.","marker":"[42]"},{"why":"Together with [53], it provides Proposition B.3 identifying the first Taylor coefficient of the twisted Kontsevich morphism as contraction by the square root of the $\\hat{A}$ cocycle.","marker":"[50]"},{"why":"Together with [50], it provides Proposition B.4 identifying the zeroth Taylor coefficient of the twisted Shoikhet morphism as multiplication by the square root of the $\\hat{A}$ cocycle.","marker":"[53]"},{"why":"Provides the divergence identity (Lemma 3.16) used to convert the $\\hat{A}$ class of the Fedosov algebroid into its Todd class.","marker":"[31]"},{"why":"Announces the formality theorem for dg manifolds whose full proof is carried out in Section 4 of this paper as Theorem B.","marker":"[32]"}],"fun_headline_variants":["Todd twist proves Duflo–Kontsevich for dg manifolds","Full calculus isomorphism on dg manifolds via Todd class","Kontsevich–Shoikhet conjecture settled by Todd twist","Todd class aligns Cartan and Tamarkin–Tsygan on dg manifolds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument's load-bearing premise is that the Fedosov dg Lie algebroid of a dg manifold is connected to the tangent dg Lie algebroid by contractions that preserve the full calculus structure (all products, brackets, contractions, derivatives, and the Connes–Rinehart operator), as asserted in an unpublished companion preprint; if those contractions preserve only part of the structure, the calculus-level isomorphism collapses.","fun_headline_variants_meta":{"raw":{"variants":["Todd twist proves Duflo–Kontsevich for dg manifolds","Full calculus isomorphism on dg manifolds via Todd class","Kontsevich–Shoikhet conjecture settled by Todd twist","Todd class aligns Cartan and Tamarkin–Tsygan on dg manifolds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000854,"raw_usage":{"total_tokens":3687,"prompt_tokens":896,"completion_tokens":2791,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":512,"completion_tokens_details":{"reasoning_tokens":2707}},"tokens_in":512,"tokens_out":2791,"duration_ms":21534,"temperature":1.0,"reasoning_tokens":2707,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T05:29:54.297718+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a dg manifold whose Atiyah cocycle is nonzero, for instance the affine space $\\mathbb{R}^{0|2}$ with a homological vector field $Q$ containing a nonzero quadratic term, and evaluate the zeroth Taylor coefficient $(\\operatorname{td}^{1/2}\\circ\\mathrm{HKR})$ on a one-polyjet $\\zeta$. If the cohomology class of $d_{\\mathrm{DR}}((\\operatorname{td}^{1/2}\\circ\\mathrm{HKR})(\\zeta))$ is not equal to the class of $(\\operatorname{td}^{1/2}\\circ\\mathrm{HKR})(B\\zeta)$, where $B$ is the Connes–Rinehart operator, then the claimed calculus isomorphism fails; the theorem predicts equality for every such $\\zeta$ and every such $Q$.","supporting_citations":[{"cited_title":"On the Duflo formula for 𝐿∞-algebras and 𝑄-manifolds","cited_arxiv_id":null,"evidence_quote":"Formulates the Kontsevich–Shoikhet conjecture, confirmed at cohomology level by Theorem A, that Todd-twisted HKR maps give a Duflo-type isomorphism for dg manifolds."},{"cited_title":"Formal geometry and Tamarkin–Tsygan calculi for dg mani- folds","cited_arxiv_id":null,"evidence_quote":"Supplies Theorem 2.5, the calculus-preserving Fedosov contractions between the tangent and Fedosov dg Lie algebroids, which let the proof descend from Fedosov calculi to dg-manifold calculi."},{"cited_title":"Deformation quantization of Poisson manifolds","cited_arxiv_id":null,"evidence_quote":"Kontsevich's formality theorem for polyvector fields to polydifferential operators on a trivialized graded manifold, used leafwise to build the Fedosov formality morphisms."},{"cited_title":"A proof of the Tsygan formality conjecture for chains","cited_arxiv_id":null,"evidence_quote":"Shoikhet's formality theorem for Hochschild chains gives the $L_\\infty$ module morphism from polyjets to differential forms whose zeroth Taylor coefficient is the twisted HKR map."},{"cited_title":"The Kontsevich weight of a wheel with spokes pointing outward","cited_arxiv_id":null,"evidence_quote":"Together with [53], it provides Proposition B.3 identifying the first Taylor coefficient of the twisted Kontsevich morphism as contraction by the square root of the $\\hat{A}$ cocycle."},{"cited_title":"A counterexample to the quantizability of modules","cited_arxiv_id":null,"evidence_quote":"Together with [50], it provides Proposition B.4 identifying the zeroth Taylor coefficient of the twisted Shoikhet morphism as multiplication by the square root of the $\\hat{A}$ cocycle."},{"cited_title":"Formality and Kontsevich-Duflo type theorems for Lie pairs","cited_arxiv_id":null,"evidence_quote":"Provides the divergence identity (Lemma 3.16) used to convert the $\\hat{A}$ class of the Fedosov algebroid into its Todd class."},{"cited_title":"Formality theorem for differential graded manifolds","cited_arxiv_id":null,"evidence_quote":"Announces the formality theorem for dg manifolds whose full proof is carried out in Section 4 of this paper as Theorem B."}],"review_version":1}