{"id":"32e9eaf6-f70b-4776-9c54-1b3f3a565819","arxiv_id":"1908.02868","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A cocycle model for complex analytic equivariant elliptic cohomology is constructed, with explicit equivariant Euler and Thom classes and a unique equivariant MString orientation.","lead":"This paper builds a new de Rham cocycle model for complex analytic equivariant elliptic cohomology with Lie group symmetries. The model yields explicit Euler and Thom classes and a unique equivariant refinement of the MString orientation.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 7.11's existence and equality claims rest on an unverified 'by inspection' descent statement and on transformation laws deferred to [BET19]; a derivation is needed.","rationale":"The reader's conditional verdict identifies the 'by inspection' uniqueness assertion in Theorem 7.11 as the weakest assumption. I agree. After reading the full argument, the real epistemic gap is that the paper offers no derivation that (70) satisfies the gluing/descent conditions defining the sheaf of equivariant elliptic cocycles on Bun_{U(1)^3}(E), and the needed transformation laws for σ and υ are imported from [BET19] rather than proved. This matters because the theorem's existence claim depends on that descent. The uniqueness claim, by contrast, is very likely automatic: a holomorphic section of a holomorphic line bundle on the universal elliptic curve stack that vanishes to infinite order along the identity section is zero, so the completion map from global sections to formal power series is injective; hence a lift, if it exists, is unique. Thus the 'unique equivariant extension' phrase, while terse, is probably justifiable. The equality with the twisted Euler class of V3 is also a matter of comparing the theta-ratio (70) with the product σ/υ classes from §6.4, which again uses the same transformation laws. So the single most load-bearing concern is the unstated verification of descent, not a suspected counterexample. A concrete independent derivation of the transformation properties and the resulting section would settle it. This does not change the reader's CONDITIONAL verdict: the paper should either supply the verification or cite it precisely.","tokens_in":48121,"tokens_out":6875,"duration_ms":77207,"concrete_test":"Independently re-derive the transformation laws of σ(τ,z) and υ(τ,z) under the lattice Z^2 and SL2(Z) from the product formulas (46) and (48), then use them to check that the expression (70) satisfies the lattice invariance, SL2(Z)-equivariance, and analyticity conditions of Definition 3.3 for U(1)^3 acting on a point, so it defines a global section of Θ3(O(-0)) on Bun_{U(1)^3}(E). Also verify that its restriction along the completion map at the identity is the non-equivariant class s of (70). If this check passes, Theorem 7.11 is established; if the descent fails, the theorem is not justified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline result, Theorem 7.11, is dispatched in two 'by inspection' sentences. Uniqueness of the equivariant refinement is claimed to follow because the non-equivariant cubical structure formulas (70) have a unique equivariant extension given by the same formulas; equality with the twisted equivariant Euler class of V3 is likewise 'on inspection.' The paper does not derive the transformation properties of σ and υ that would be needed to verify that (70) satisfies the descent data of Definition 3.3 for the sheaf on Bun_{U(1)^3}(E), nor does it verify that the completion map along the zero section is injective on global sections of Θ3(O(-0)). It also relies on the companion preprint [BET19] for the transformation laws of the sigma and upsilon functions (Definition 6.5, Proposition 6.10), so the existence of the equivariant lift is not self-contained. The load-bearing step is therefore not just a missing formula but the verification that the formula (70) genuinely defines a global section of the twisted sheaf on the stack; without it, the existence of any equivariant refinement is an assertion. This is a correctness-risk concern, not a fatal one: the sigma-function identities are standard, and the proof is likely recoverable.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs a sheaf of commutative differential graded algebras, denoted Ell_G(M), on the moduli stack Bun_G(E) of G-bundles over elliptic curves, with local sections given by compatible equivariant de Rham complexes on fixed-point loci. It claims that restricting to a fixed elliptic curve recovers Grojnowski's complex analytic equivariant elliptic cohomology for connected G, and that for finite G the global sections recover Devoto's theory. It then constructs explicit Mathai-Quillen type Euler and Thom classes using products of sigma functions, identifies the relevant twistings with Looijenga line bundles and level-1 loop-group characters, and gives an elliptic Chern-Weil map. The final section studies equivariant refinements of orientations: it shows that ordinary complex orientations do not refine equivariantly, constructs a twisted refinement via the function upsilon, and claims in Theorem 7.11 that there is a unique equivariant refinement of the MO<8>-orientation, equal to the twisted equivariant Euler class of the virtual bundle V3=(L1-1) otimes (L2-1) otimes (L3-1), obtained from the theorem-of-the-cube formula (70).","tokens_in":48392,"tokens_out":10039,"duration_ms":118040,"significance":"If the claims hold, the paper provides a useful and explicit differential-geometric model for complex analytic equivariant elliptic cohomology: it gives a uniform cocycle-level framework, recovers Grojnowski's and Devoto's theories, and produces concrete representatives of elliptic Euler and Thom classes tied to loop-group characters and the Atiyah-Segal completion perspective. The stack Bun_G(E), the holomorphic structure on it, and the treatment of twisting line bundles are carefully set up, and the examples involving U(1)-actions on spheres are informative. However, the headline uniqueness theorem is currently asserted rather than proved, and several load-bearing transformation laws are imported from the companion preprint [BET19], so the central statement needs substantial repair before the results can be regarded as fully established.","major_comments":[{"comment":"The proof of Theorem 7.11 is not sufficient for the paper's headline claim. The assertion that formulas (70) 'by inspection' have a unique equivariant extension skips the verification that the displayed function defines a global section of Θ3(O(-0)) on (E∨)^3, or equivalently on Bun_{U(1)^3}(E): one must check quasi-periodicity in each variable under the cocharacter lattice, the rigid, symmetric, and cocycle conditions, and the SL2(Z)-equivariance and descent data required by Definition 3.3. The uniqueness statement is also asserted without proof: the paper does not show that the completion map from global sections of Θ3(O(-0)) to the formal power series ring is injective, nor does it otherwise rule out other equivariant sections with the same non-equivariant expansion. Since Theorem 7.11 is the central result advertised in the abstract, this step needs a real derivation rather than an inspection claim.","section":"§7.3, Theorem 7.11"},{"comment":"The construction of the line bundles A_G and L_G and their identification with level-1 Looijenga line bundles depends on transformation formulas for σ and υ quoted as [BET19, Equations 146-149] and not reproduced in this paper. These transformation laws are needed for Proposition 6.8, for the gluing of the Thom forms in Proposition 6.13, and ultimately for the twisted equivariant Euler class used in Theorem 7.11(2). The paper should either include these formulas or prove them directly; delegating a load-bearing step to a companion preprint is not adequate for the claims made here.","section":"§6.4, Definition 6.5 and Proposition 6.10"},{"comment":"The proof of Proposition 3.5(1) is a sketch of the key descent construction. It asserts that the data assemble via Proposition 2.14 and that conditions (C1)-(C2) follow from the invariance property (25), but it does not actually verify the equivariance of the restriction maps (28) under composition, the independence of choices of logarithms and maximal commuting subalgebras, or the sheaf condition on arbitrary open covers. Since this proposition defines the central object of the paper, the argument should be written out in enough detail to be checked.","section":"§3, Proposition 3.5"},{"comment":"The proof of Theorem 4.1 is compressed at a technically sensitive point: the passage from W-invariants of the sum over W/W_h to W_h-invariants of the identity summand is stated without justification, and the final stalk argument is asserted to be clear. This is not necessarily wrong, and the comparison is plausible, but given that the recovery of Grojnowski's theory is one of the paper's main advertised results, a short proof or a precise reference for this finite-group induction step should be supplied.","section":"§4, Theorem 4.1"}],"minor_comments":[{"comment":"The phrase \"can be identifies\" should read \"can be identified.\"","section":"§4.1, Definition 4.2"},{"comment":"The letter E is used both for the universal elliptic curve and for the restricted family over U⊂H; this notational overlap should be resolved for clarity.","section":"§7.1, Definition 7.6"},{"comment":"The phrase \"Wh-invariant elements of cdga associated with H×T^ε_h\" is awkward; it would be clearer to say \"Wh-invariant sections of the sheaf of cdgas on H×T^ε_h.\"","section":"§3, proof of Proposition 3.5"},{"comment":"The definition of an equivariant refinement would be clearer if it explicitly stated that the class is a global section of the Θ3(O(-0))-twisted sheaf on Bun_{U(1)^3}(E), rather than leaving this to the proof.","section":"§7.3, Definition 7.10"}],"recommendation":"major_revision","confidential_remarks":"The companion preprint [BET19] is cited for essential technical formulas, so the editor should confirm that this companion is available and stable. The central uniqueness theorem is currently not proved to the standard claimed; a revision should expand the proof substantially rather than rely on 'by inspection' assertions."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Send this to a serious referee. It does something real: a sheaf of cdgas on Bun_G(E) that refines Grojnowski for connected groups and Devoto for finite groups, with explicit Mathai–Quillen type Euler and Thom cocycles built from sigma functions. The comparison theorems in Sections 4 and 5 are clearly formulated and plausible; the maps are natural and I did not find an obvious gap in the main construction. The connection between the Euler classes and level 1 loop group characters is a genuine payoff, and the worked examples give a useful computational handle.\n\nThe soft spot is exactly where the stress-test note points. Theorem 7.11, the headline uniqueness claim for the equivariant MString orientation, is dispatched as 'by inspection' without a derivation of the equivariant extension of the cubical structure (70). The needed transformation laws for sigma and upsilon are deferred to the companion [BET19], and the descent to Bun_{U(1)^3}(E) is asserted rather than checked. This is a correctness-risk, not a fatal flaw: the sigma identities are standard and the proof is likely recoverable, but as written the existence part of the uniqueness theorem is an assertion. Proposition 3.5 is also a sketch, though that piece looks more routine and less load-bearing.\n\nThe citation pattern is honest; reliance on [BET19] is a companion-paper dependency, not a flaw, and there is no circularity in the main comparisons. The paper deserves referee time and will probably become a standard reference for this cocycle model. My recommendation: conditional accept, with the referee instructed to insist on a real proof of Theorem 7.11 and either reproduce or clearly state the imported transformation laws.","headline":"A genuinely new cocycle model for complex analytic equivariant elliptic cohomology, with explicit Euler/Thom cocycles from sigma functions; the headline uniqueness claim rests on a 'by inspection' proof and deferred transformation laws, so send to a serious referee but demand the missing derivation.","tokens_in":48878,"tokens_out":2366,"would_cite":true,"duration_ms":26811,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["55N34","55N91","14H52","22E67"],"pacs":[],"model":"deepseek-v4-flash","headline":"Complex analytic equivariant elliptic cohomology admits a unique refinement of the string orientation: the twisted equivariant Euler class of V3, realized inside a sheaf of differential graded algebras on the moduli space of G-bundles…","keywords":["equivariant elliptic cohomology","cocycle model","de Rham model","string orientation","theorem of the cube","Looijenga line bundles","loop group representations","Weierstrass sigma function"],"falsifier":"Compute the space of $\\mathrm{SL}_2(\\mathbb{Z})$-equivariant holomorphic sections of the cubical line bundle $\\Theta_3(\\mathcal{O}(-0))$ over the universal triple $E^{\\vee}\\times E^{\\vee}\\times E^{\\vee}$: Theorem 7.11 predicts this space is one-dimensional, spanned by formula (70). Exhibiting a second linearly independent section, for instance one differing from (70) by a nonconstant modular function or by a character of the elliptic curve, would falsify the claimed uniqueness; so would a direct failure of the rigid, symmetric, or cocycle conditions for any putative alternative section at the stalk level.","tokens_in":47959,"feed_emoji":"➰","tokens_out":17934,"duration_ms":162487,"temperature":0.7,"pith_summary":"Equivariant elliptic cohomology is meant to be the arena for elliptic character theory, where loop-group representations and the geometry of elliptic curves meet, but the complex-analytic versions that existed came in two separate regimes: Grojnowski's theory for connected groups and Devoto's for finite groups. This paper claims both are facets of a single object: a sheaf of commutative differential graded algebras on the stack $\\mathrm{Bun}_G(E)$ of flat $G$-bundles over elliptic curves, whose local data are equivariant differential forms on fixed-point manifolds $M^{h}$ for commuting pairs $h=(h_1,h_2)$. The model yields explicit equivariant Euler and Thom cocycles from the Weierstrass $\\sigma$ function, identified with level-one vacuum characters of loop groups. The paper's central result is that the string (MString) orientation of complex analytic elliptic cohomology, built from the theorem of the cube, has exactly one equivariant refinement, and that refinement is the twisted equivariant Euler class of the universal virtual bundle $V_3$.","feed_headline":"One formula gives the string orientation a unique equivariant lift","feed_subtitle":"A sheaf of differential forms on G-bundles over elliptic curves unifies two prior theories and fixes the MString class.","key_machinery":"The object that carries the whole construction is the sheaf of elliptic cocycles, $\\widehat{\\mathrm{Ell}}_G^{\\bullet}(M)$: a local section assigns to every commuting pair $h=(h_1,h_2)$ in a neighborhood in $\\mathrm{Bun}_G(E)$ a $G^{h}_0$-equivariant differential form on the fixed-point manifold $M^{h}$, subject to a conjugation-invariance condition and an analyticity condition, $\\alpha_{h'}(X)=\\operatorname{res}\\alpha_h(X+(X_1-\\tau X_2))$, that lets the forms vary holomorphically as $h$ is deformed. The specific formula carrying the orientation result is the cubical structure of the line bundle $\\mathcal{O}(-0)$ on the elliptic curve, the unique section, by the theorem of the cube, of $\\Theta_3(\\mathcal{O}(-0))$ on $E\\times E\\times E$, written in $\\sigma$-function coordinates as $$ s=\\frac{\\$\\sigma$(x+y)\\$\\sigma$(x+z)\\$\\sigma$(y+z)\\$\\sigma$(0)}{\\$\\sigma$(x+y+z)\\$\\sigma$(x)\\$\\sigma$(y)\\$\\sigma$(z)} , $$ which is formula (70) of the paper. The proof of uniqueness asserts that this same formula is the unique equivariant extension, and matches it with the twisted equivariant Euler class of $V_3$. The $\\sigma$ functions $\\sigma(\\tau,z)$ and $\\upsilon(\\tau,z)$ also supply the Euler and Thom cocycles for $U(n)$ and $\\mathrm{Spin}(2n)$, linking them to loop-group characters.","core_discovery":"On the paper's own terms, the discovery is that complex analytic equivariant elliptic cohomology is carried by a single sheaf of cdgas: for a $G$-manifold $M$, the sheaf $\\widehat{\\mathrm{Ell}}_{G}^{\\bullet}(M)$ on $\\mathrm{Bun}_G(E)$ whose local section at a commuting pair $h=(h_1,h_2)$ is a $G^{h}_0$-equivariant differential form on the fixed-point manifold $M^{h}$, satisfying conjugation invariance and an analyticity condition, $\\alpha_{h'}(X)=\\operatorname{res}\\alpha_h(X+(X_1-\\tau X_2))$, that encodes the additive structure of the elliptic curve. Restricting the cohomology sheaf to a fixed elliptic curve and a maximal torus reproduces Grojnowski's connected-group theory (Theorem 4.1), and for finite groups the global sections reproduce Devoto's theory over $\\mathbb{C}$ (Theorem 5.1). The paper then constructs Mathai–Quillen type equivariant elliptic Euler and Thom classes for $U(n)$ and $\\mathrm{Spin}(2n)$ from $\\theta$ functions, identifies them with supercharacters of level-one vacuum representations of loop groups, and proves that the $MO\\langle 8\\rangle$ (MString) orientation of complex analytic elliptic cohomology has a unique equivariant refinement, equal to the twisted equivariant Euler class of $V_3=(L_1-1)\\otimes(L_2-1)\\otimes(L_3-1)$; uniqueness is a consequence of the theorem of the cube for the line bundle $\\mathcal{O}(-0)$.","pith_inferences":["If the uniqueness claim holds, a natural expectation is that the same equivariant rigidity applies to any elliptic cohomology theory over $\\mathbb{C}$ whose formal group carries the standard coordinate: the phenomenon would be a feature of elliptic curves themselves, not of this particular model.","The paper leaves open the derived global sections of the sheaf for general $G$; computing them for examples such as $U(1)$ acting on $S^2$ would test whether the cocycle model is the right input for elliptic Springer theory and elliptic stable envelope constructions.","The asserted 'by inspection' uniqueness in the proof of Theorem 7.11 deserves a direct check: computing the space of $\\mathrm{SL}_2(\\mathbb{Z})$-equivariant holomorphic sections of $\\Theta_3(\\mathcal{O}(-0))$ over the universal family would either confirm the rigidity or expose hidden moduli of equivariant orientations.","The identification of Thom cocycles with loop-group characters suggests an elliptic index theorem: pushing forward equivariant elliptic operators along these Thom classes should yield modular forms whose $q$-expansions are governed by the corresponding characters, a testable analogue of the index-theoretic reading of equivariant K-theory."],"forward_implications":["One construction covers both regimes: connected groups recover the earlier cocycle-level theory and finite groups recover the earlier finite-group theory, so results proved in one formalism now transfer to the other.","The Euler and Thom cocycles give explicit, computable representatives for the $MU\\langle 6\\rangle$- and MString-orientations of complex analytic elliptic cohomology, upgrading those orientations from existence statements to formulas.","Because the Euler cocycles are supercharacters of level-one vacuum representations, characteristic classes in this theory come with a representation-theoretic reading: loop-group characters are the new Chern roots.","Complex orientations of elliptic cohomology admit no equivariant refinement at all, so the twisted equivariant setting is not a convenience but a necessity, and the uniqueness from the theorem of the cube removes the ambiguities that exist at the level of formal groups.","Global sections of the sheaf are 24-periodic rather than 2-periodic, and the theory's derived global sections are nontrivial, matching known computations of equivariant topological modular forms for $U(1)$."],"supporting_citations":[{"why":"The connected-group equivariant elliptic cohomology theory that the paper's sheaf refines into a cocycle model (Theorem 4.1).","marker":"[Gro07]"},{"why":"Provides the definition, adapted in Section 5.1, of the finite-group equivariant elliptic cohomology that the model reproduces (Theorem 5.1).","marker":"[Dev98]"},{"why":"The theorem-of-the-cube construction of the MString-orientation whose unique equivariant refinement is the subject of Theorem 7.11.","marker":"[Hop94]"},{"why":"Shows how cubical structures produce the $MU\\langle 6\\rangle$ and MString orientations and supplies the characteristic class being refined in equation (74).","marker":"[AHS01]"},{"why":"Supplies Lemma 1.3 controlling fixed-point sets under deformations of commuting pairs, and the de Rham bouquet model that Definition 3.3 generalizes.","marker":"[BG94]"},{"why":"Companion paper from which the transformation laws for the sigma-function line bundles (Definition 6.5, Proposition 6.10) are imported.","marker":"[BET19]"},{"why":"Loop-group representation characters (Corollary 10.9) that identify sections of Looijenga line bundles with the Verlinde algebra in Proposition 3.17.","marker":"[And00]"},{"why":"The identity expressing the Weierstrass sigma function as the normalized vacuum character, used to write the Euler and Thom cocycles.","marker":"[Zag86]"},{"why":"The Mathai–Quillen construction of equivariant Thom forms that the elliptic Thom cocycles of Definition 6.11 are built from.","marker":"[MQ86]"},{"why":"Constructs the line bundles on the commuting-pair space used as twists for finite groups, tying the finite-group theory to Chern–Simons theory.","marker":"[FQ93]"}],"fun_headline_variants":["De Rham sheaf unifies equivariant elliptic cohomology and lifts MString","Unique equivariant MString lift from one de Rham cocycle model","Theta-twisted Euler class fixes equivariant elliptic orientation","Cocycle model merges Grojnowski and Devoto, refines MString","Equivariant elliptic de Rham model sharpens MString orientation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The headline uniqueness theorem rests on the proof's assertion, made by inspection without derivation, that the non-equivariant cubical structure formula (70) has exactly one equivariant extension given by the same formula; the construction also imports sigma-function transformation laws from the companion preprint [BET19] without reproducing them.","fun_headline_variants_meta":{"raw":{"variants":["De Rham sheaf unifies equivariant elliptic cohomology and lifts MString","Unique equivariant MString lift from one de Rham cocycle model","Theta-twisted Euler class fixes equivariant elliptic orientation","Cocycle model merges Grojnowski and Devoto, refines MString","Equivariant elliptic de Rham model sharpens MString orientation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000853,"raw_usage":{"total_tokens":3725,"prompt_tokens":978,"completion_tokens":2747,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":594,"completion_tokens_details":{"reasoning_tokens":2648}},"tokens_in":594,"tokens_out":2747,"duration_ms":19425,"temperature":1.0,"reasoning_tokens":2648,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:30:48.917519+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the space of $\\mathrm{SL}_2(\\mathbb{Z})$-equivariant holomorphic sections of the cubical line bundle $\\Theta_3(\\mathcal{O}(-0))$ over the universal triple $E^{\\vee}\\times E^{\\vee}\\times E^{\\vee}$: Theorem 7.11 predicts this space is one-dimensional, spanned by formula (70). Exhibiting a second linearly independent section, for instance one differing from (70) by a nonconstant modular function or by a character of the elliptic curve, would falsify the claimed uniqueness; so would a direct failure of the rigid, symmetric, or cocycle conditions for any putative alternative section at the stalk level.","supporting_citations":[],"review_version":1}