{"id":"fc44da9c-99bb-4988-876f-674647bfdea7","arxiv_id":"2411.09401","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Hermitian K-theory of an exact form category is canonically equivalent to the hermitian K-theory of its derived Poincaré ∞-category, with no assumption that 2 is invertible.","lead":"This paper proves that two different mathematical settings for hermitian K-theory compute the same invariant in full generality. The result matters because it lets mathematicians use modern higher-category tools without sacrificing the older chain-complex methods, even when the number 2 is not invertible.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem B's spectrum-level reduction to complicial categories rests on an unstated spectrum-level Gillet-Waldhausen theorem; the quoted Theorem 2.2.1 is only space-level.","rationale":"The reader identified Theorem 2.2.1 as the weakest assumption, and I agree that the Gillet-Waldhausen bridge is load-bearing. My concern sharpens this: Theorem 2.2.1 is quoted as a space-level equivalence, while Theorem B needs the same comparison at the level of spectra. Theorem 4.3.5 concerns only complicial categories, so the reduction from exact categories to complicial ones at spectrum level is precisely the missing stated ingredient. The gap may be fillable from [Sch24b], which is why I do not recommend changing the conditional verdict. The paper is otherwise careful: the comparison of Grothendieck-Witt spaces in Sections 3.2–3.4, the explicit model for the derived quadratic functor in Section 2, and the delooping compatibility argument in Section 4.3 are internally detailed. The concern is therefore about an unstated external input, not an evident internal contradiction.","tokens_in":64110,"tokens_out":7850,"duration_ms":75285,"concrete_test":"Check [Sch24b], especially §11 near Definition 11.5, for a spectrum-level Gillet-Waldhausen theorem. If it exists, cite it explicitly in the proof of Theorem B. If it does not, verify directly at the first delooping level: compare the positive Ω-spectra of Definition 4.1.1 by checking that for each n ≥ 0, the map wQuad(R^{(n)}(E[n], Q[n], iso)) → wQuad(R^{(n)}(Ch^b(E)[n], Q[n], qis)) is a weak equivalence compatible with the bonding maps γ_n of [Sch24b] and δ_n of [CDH+II]. The n = 1 commutativity check is the minimal test: if the diagram comparing the two delooping maps fails to commute, Theorem B's spectrum-level assertion is not established by the text.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central spectrum-level claim for arbitrary exact form categories is Theorem B: GW(E,Q) ≃ GW(D^b(E), Ϙ). The proof route is to embed E into Ch^b(E), apply Theorem 4.3.5 to the complicial category (Ch^b(E), Q^ch, qis), and then identify GW(E,Q) with GW(Ch^b(E), Q^ch, qis). The only quoted identification is Theorem 2.2.1, i.e. [Sch24b, Thm 10.5], which is explicitly stated as a weak equivalence of Grothendieck-Witt spaces, not spectra. Section 4 defines GW spectra for complicial categories in Definition 4.1.1 and proves the spectrum-level comparison only for complicial inputs in Theorem 4.3.5. No spectrum-level version of the Gillet-Waldhausen comparison is stated or proved in this paper; Corollary 3.4.2 is explicitly space-level. Consequently, the passage from an arbitrary exact form category to its bounded chain-complex category is an unstated load-bearing assumption at the spectrum level. If [Sch24b] contains such a spectrum-level theorem (e.g. in its §11), the omission is expository; if not, Theorem B does not follow from the arguments presented. A similar external reliance occurs in §5.1, where Lemma 5.1.2 invokes weak idempotent completeness via [Sch24b, Lem 10.6] without proving the reduction.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper claims a canonical equivalence between the hermitian K-theory (Grothendieck-Witt) spectrum of an exact form category in the sense of Schlichting and that of its derived Poincaré infinity-category in the sense of Calmes-Dotto-Harpaz-Hebestreit-Land-Moi-Nardin-Nikolaus-Steimle, without assuming invertibility of 2. The main result, Theorem 4.3.5, states that for a complicial exact form category with weak equivalences and strong duality, the localization functor induces an equivalence of Grothendieck-Witt spectra. The proof proceeds by constructing explicit simplicial models for the nonabelian derived functor of the quadratic functor, comparing the hermitian S-dot and Q-dot constructions, and then identifying the deloopings at spectrum level. The introduction also states a spectrum-level Theorem B for arbitrary exact form categories via the embedding into bounded chain complexes, and a corollary about genuine symmetric Poincare structures.","tokens_in":64401,"tokens_out":15661,"duration_ms":151870,"significance":"If the advertised comparisons hold, the paper closes an important gap between the classical 1-categorical hermitian K-theory of exact categories and the infinity-categorical Poincare formalism, with no restriction on the prime 2. The proof strategy is original and detailed: the explicit model for the nonabelian derived functor, the use of complicial exact categories as categories of fibrant objects, and the systematic comparison of the two S-dot constructions are valuable technical contributions. The paper is carefully structured and contains many explicit lemmas that make the main argument, for complicial inputs, largely checkable. Its main weakness is that a load-bearing spectrum-level Gillet-Waldhausen comparison needed for Theorem B is not stated or proved in the manuscript, and one auxiliary proof in Section 5.3 contains an assertion that appears to need substantial justification.","major_comments":[{"comment":"The spectrum-level Theorem B stated in the introduction is not proved in the main text. Theorem 2.2.1, quoted from [Sch24b, Thm 10.5], is explicitly a weak equivalence of Grothendieck-Witt spaces, and Corollary 3.4.2 is likewise only a space-level comparison. Definition 4.1.1 constructs the Grothendieck-Witt spectrum only for complicial exact form categories with weak equivalences, and Theorem 4.3.5 proves the spectrum-level comparison only for such complicial inputs. The passage GW(E,Q,iso) ≃ GW(Ch^b(E),Q^ch,qis) needed to reduce Theorem B to Theorem 4.3.5 is therefore an unstated spectrum-level Gillet-Waldhausen theorem. Please either prove this spectrum-level comparison, state it as an explicit theorem with proof, or cite precisely a spectrum-level result in [Sch24b] and explain how it applies. The same issue affects the reduction to weak idempotent completion in §5.1, which invokes [Sch24b, Lem 10.6] at the space level.","section":"§4.3, Introduction (Theorem B), Corollary 3.4.2"},{"comment":"The proof of Lemma 5.3.1 asserts that hom_{K^b(E)}(x,D(x))^{hC2} is coconnective for x in the heart. For a discrete spectrum with a C2-action, homotopy fixed points generally have nontrivial positive homotopy groups; for example, a trivial C2-action on an Eilenberg-Mac Lane spectrum HZ gives H^{2k}(BC2;Z)=Z/2 for k≥1. This assertion therefore needs a justification that is not supplied. Since the conclusion that Ϙ_{≥0}(x) is concentrated in degree 0 is used directly in the proof of Lemma 5.3.1 and hence in the corollary on genuine symmetric structures, this step is load-bearing for that part of the paper.","section":"§5.3, Lemma 5.3.1"}],"minor_comments":[{"comment":"The notation Q^{∆^1} for the form functor on the arrow category and Q^{∆^●} for the simplicial derived functor are nearly identical and are used close together in the bonding-map argument. The warning in the text is helpful, but the notation should be changed or consistently distinguished to avoid serious reader confusion.","section":"§4.1, §4.3"},{"comment":"The introduction states Theorem A for a 'complicial exact form category with weak equivalences' without explicitly saying 'with strong duality', although strong duality is used essentially in Theorem 2.5.10 and throughout Section 3. Please make the hypothesis explicit in the theorem statement.","section":"Introduction, Theorem A"},{"comment":"The manuscript contains several typos and OCR-style artifacts, for example 'exact from category' in §4.1 and 'we likeqwise construct' in Remark 2.1.8. A proofreading pass would improve readability.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The central spectrum-level comparison for complicial categories (Theorem 4.3.5) appears to be well supported by the arguments presented, assuming the quoted additivity and comparison results from the companion preprint [Sch24b]. The main gap is the unproved spectrum-level Gillet-Waldhausen comparison needed for Theorem B; if [Sch24b] contains this theorem, the authors should cite it explicitly, and if not, they need to prove it. The coconnectivity assertion in Lemma 5.3.1 also deserves close scrutiny before publication. These issues are fixable within the manuscript's scope, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Main take: this paper proves a genuine spectrum-level comparison between Schlichting's form-categorical hermitian K-theory and the Poincaré ∞-category formalism, for complicial exact form categories with weak equivalences, with no assumption on 2. Theorem A (4.3.5) is the real new result, and the proof is detailed and serious throughout. The advertised Theorem B for arbitrary exact form categories, however, has a gap in the written argument: the reduction from an exact form category E to its bounded chain complex category is only justified at the level of Grothendieck-Witt spaces (Theorem 2.2.1, Corollary 3.4.2), while the spectrum-level comparison in Section 4 is only proved for complicial inputs. The stress-test note is right. Unless [Sch24b] already contains an explicit spectrum-level Gillet-Waldhausen theorem that the authors simply omitted to cite, Theorem B does not follow from the arguments as presented. The missing statement is exactly what a referee should ask for, and it is likely repairable, but it should not be left implicit.\n\nWhat the paper does well: it gives a clean model for the nonabelian derived functor of a quadratic functor on a complicial exact category, proves the derived Poincaré structure is genuinely Poincaré, and carries the space-level comparison through with care. The appendices on complicial exact categories and polynomial functors are useful and not just filler. The paper is also honest about its reliance on [Sch24b] and the unpublished [CDH+II] series. That reliance is heavy, but it is standard in this area and the authors clearly know which inputs they are importing.\n\nThe main soft spot is the spectrum-level reduction in Theorem B. A secondary issue is that the paper leans on unpublished work for several load-bearing structural results, which makes verification slower but not impossible. I do not see any circularity or fitted claims; the central equivalence is genuinely derived from prior frameworks.\n\nWho this is for: researchers in hermitian K-theory, motivic homotopy theory, and higher algebra. It deserves a serious referee, not a desk reject. I would send it out with a specific request to check the spectrum-level Gillet-Waldhausen claim and to make the proof of Theorem B explicit.","headline":"Solid, important spectrum-level comparison for complicial exact form categories, but Theorem B has a real gap: the reduction to chain complexes is only proved at space level.","tokens_in":64935,"tokens_out":3859,"would_cite":false,"duration_ms":39049,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["18F25","19G12","18N60"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that passing from an exact category with forms to its derived ∞-category preserves the Grothendieck–Witt spectrum, with no assumption on 2 being invertible.","keywords":["Grothendieck-Witt groups","hermitian K-theory","exact form categories","Poincaré ∞-categories","derived categories","quadratic functors","complicial exact categories","localisation"],"falsifier":"Compute the Grothendieck–Witt spectra $\\mathrm{GW}(E,\\mathcal{Q})$ and $\\mathrm{GW}(D^b(E), \\vartheta)$ for a non-split exact form category over a field of characteristic 2, such as a category of vector bundles with a symmetric forms functor; the theorem predicts the canonical map is a weak equivalence, so any homotopy-group difference after group completion would falsify it.","tokens_in":63900,"feed_emoji":"🔄","tokens_out":9058,"duration_ms":75372,"temperature":0.7,"pith_summary":"This paper establishes that hermitian K-theory, equivalently Grothendieck–Witt theory, is invariant under localisation from an exact category with forms to its derived ∞-category. The main theorem states that for any complicial exact form category with weak equivalences, the localisation functor induces an equivalence of Grothendieck–Witt spectra between the classical 1-categorical construction and the modern Poincaré ∞-categorical construction. This supplies the missing comparison between the two frameworks, and it works without assuming that 2 is invertible, so it covers rings and schemes of characteristic 2. As a direct consequence, the Grothendieck–Witt groups of an exact category with strong duality agree with those of its bounded derived category.","feed_headline":"Localisation preserves Grothendieck-Witt spectra","feed_subtitle":"New equivalence matches 1-categorical and ∞-categorical hermitian K-theory without requiring 2 to be invertible.","key_machinery":"The load-bearing construction is the right derived quadratic functor $R\\mathcal{Q} = \\pi_! P_2 \\gamma_! \\mathcal{Q}^{\\Delta^\\bullet}$ on the localised ∞-category $L_w(E)$, where $\\mathcal{Q}^{\\Delta^\\bullet}(x)$ is the simplicial abelian group $\\mathcal{Q}(\\Delta^\\bullet \\otimes x)$ and $P_2$ denotes the 2-excisive approximation. A complicial exact form category is an exact category with forms equipped with a compatible action of the category of bounded chain complexes of free abelian groups, an action that guarantees the localisation is a stable ∞-category. The proofs identify the polarisation of this functor with the mapping spectrum $\\mathrm{map}_{L_{\\mathrm{Frob}}(E)}(x, D(y))$, which makes $(L_w(E), R\\mathcal{Q})$ a Poincaré ∞-category, and show that the quadratic functor agrees, after truncation, with the colimit formula $x \\mapsto \\mathrm{colim}_{y \\to x} \\mathcal{Q}(y)$ over trivial deflations. This identification is what allows the classical Grothendieck–Witt space, built from nondegenerate quadratic spaces, to be compared levelwise with the Poincaré-categorical space of nondegenerate forms.","core_discovery":"On the paper's own terms, the central discovery is Theorem A: for a complicial exact form category with weak equivalences $(E,\\mathcal{Q},D,\\eta,w)$, the ∞-categorical localisation $E \\to L_w(E)$ induces a functorial equivalence of hermitian K-theory spectra $\\mathrm{GW}(E,\\mathcal{Q},w) \\simeq \\mathrm{GW}(L_w(E), R\\mathcal{Q})$, where the right-hand side is the Poincaré-categorical Grothendieck–Witt spectrum of the derived Poincaré category. Theorem B specialises this to the canonical embedding $E \\hookrightarrow D^b(E)$ of an exact form category into its bounded derived ∞-category, yielding $\\mathrm{GW}(E,\\mathcal{Q}) \\simeq \\mathrm{GW}(D^b(E), \\vartheta)$. The proof exhibits the right-hand structure $R\\mathcal{Q}$ explicitly as the 2-excisive approximation of the derived presheaf $\\mathcal{Q}^{\\Delta^\\bullet}$, and uses this model to compare the classical and Poincaré Grothendieck–Witt spaces levelwise along their simplicial S-constructions before delooping to spectra.","pith_inferences":["Because the derived quadratic functor is described by a colimit formula, the same mechanism should compute derived functors for other quadratic structures, such as oriented or higher-degree forms, and may yield analogous comparison statements for Witt or L-theory.","The spectrum-level comparison suggests that the universal additive invariant of Poincaré ∞-categories is determined by its value on exact form categories, so a single localising invariant controls both frameworks.","The absence of any condition on 2 raises the possibility that derived-invariance results for hermitian K-theory can be extended to characteristic 2 without the usual polarisation caveat."],"forward_implications":["For an exact form category with strong duality, the Grothendieck–Witt spectrum of the category is equivalent to that of its bounded derived ∞-category, generalising the previously known additive and split-exact cases.","The genuine symmetric Grothendieck–Witt theory of the derived ∞-category coincides with the 1-categorical symmetric theory of the original exact category.","All comparisons hold with no assumption on the invertibility of 2, so the results apply to rings, schemes, and exact categories of characteristic 2.","The proof supplies an explicit model for the nonabelian derived functor of a nondegenerate quadratic functor on an exact category."],"supporting_citations":[{"why":"Supplies the chain-complex comparison theorem quoted as Theorem 2.2.1, the reduction that forces the whole argument.","marker":"[Sch24b]"},{"why":"Establishes the Poincaré ∞-category formalism and the ∞-categorical Grothendieck–Witt spectrum that the paper compares against.","marker":"[CDH+I]"},{"why":"Defines exact form categories and the 1-categorical Grothendieck–Witt space and spectra whose invariance is being proved.","marker":"[Sch21]"},{"why":"Provides the Poincaré-categorical Q-construction, cobordism categories, and metabolic sequences used for the spectrum-level comparison and delooping.","marker":"[CDH+II]"},{"why":"Supplies the uniqueness theorem for quadratic (2-excisive) extensions used to identify the derived quadratic functor on the derived category.","marker":"[BGMN22]"},{"why":"Provides the ∞-categorical fibrant-object and localisation machinery used to show complicial exact categories localise to stable ∞-categories.","marker":"[Cis19]"},{"why":"Gives the genuine symmetric Poincaré structures and orientations used in Section 5 to identify the derived structures.","marker":"[CHN24]"}],"fun_headline_variants":["Hermitian K-theory survives localisation","Grothendieck-Witt spectra match across derived categories","No 2-invertibility needed for hermitian K-theory equivalence","Derived Poincaré categories preserve Grothendieck-Witt","Forms K-theory: exact vs derived equivalence"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof rests on the quoted chain-complex comparison theorem (Theorem 2.2.1) that the exact embedding of an exact form category into its bounded chain-complex category induces a weak equivalence of Grothendieck–Witt spaces; if that theorem fails for some input, the comparison for general exact categories breaks down.","fun_headline_variants_meta":{"raw":{"variants":["Hermitian K-theory survives localisation","Grothendieck-Witt spectra match across derived categories","No 2-invertibility needed for hermitian K-theory equivalence","Derived Poincaré categories preserve Grothendieck-Witt","Forms K-theory: exact vs derived equivalence"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000152,"raw_usage":{"total_tokens":1156,"prompt_tokens":853,"completion_tokens":303,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":469,"completion_tokens_details":{"reasoning_tokens":221}},"tokens_in":469,"tokens_out":303,"duration_ms":3331,"temperature":1.0,"reasoning_tokens":221,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T20:39:53.578519+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the Grothendieck–Witt spectra $\\mathrm{GW}(E,\\mathcal{Q})$ and $\\mathrm{GW}(D^b(E), \\vartheta)$ for a non-split exact form category over a field of characteristic 2, such as a category of vector bundles with a symmetric forms functor; the theorem predicts the canonical map is a weak equivalence, so any homotopy-group difference after group completion would falsify it.","supporting_citations":[],"review_version":1}