{"id":"db100091-3f1c-45f4-af87-5ef8ac1f36c9","arxiv_id":"2412.03231","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper proves an ∞-categorical extension theorem (Theorems A and B) that extends functors on grids of commutative squares to compactifications, a load-bearing step in constructing exceptional pushforwards and the abstract six-functor formalism.","lead":"This mathematics paper proves a technical gluing theorem used to build six-functor formalisms, the abstract framework behind modern cohomology theories in algebraic and arithmetic geometry. It re-derives, with new combinatorial machinery, results first obtained by Liu and Zheng in unpublished notes.","discovery_kind":"replication","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Proof of Prop 4.1.8 takes limits in CE2,/y along E1-labeled diagrams; weak contractibility of Kpt(τ) is unsupported as written.","rationale":"The reader's weakest_assumption correctly identifies Proposition 4.1.8 as the most load-bearing unresolved point. The central claim Theorem 1.0.3 is proven by combining Theorem 4.2.1 (extension along pcomm) and Theorem 5.2.1 (extension along pcart). Theorem A's proof hinges on the weak contractibility of Kpt(τ), which is exactly Proposition 4.1.8. The internal inconsistency between Definition 4.1.7.4 (E1-pointwise morphisms) and the use of CE2,/y limits in Lemma 4.1.9 is a concrete, non-cosmetic gap: if the diagram edges are E1, the limit in CE2,/y does not apply, and without CE1,/y finite limits the cofilteredness argument collapses. This is independent of the additional, equally real issue that Theorem 5.2.1 uses Proposition 5.1.9, whose pullback hypothesis is absent from the theorem statement. Both concerns point to the same systemic problem: the stated hypotheses of Theorems 1.0.3, 4.2.1, and 5.2.1 do not include the pullback-completeness on which the proofs rely. However, the gap is repairable—by adding the missing hypotheses or by reworking the edge-class bookkeeping—so the manuscript is best treated as conditional rather than rejected. The reader's CONDITIONAL verdict with MODERATE confidence remains appropriate; no change is needed.","tokens_in":28366,"tokens_out":9261,"duration_ms":79268,"concrete_test":"Re-derive Lemma 4.1.9 while explicitly tracking edge classes: take the map gm: ∂∆^m → Kpt(τ), which by Definition 4.1.7.4 has edges in E1. Write out the induced diagram in C/y and check whether its edges belong to E2 or E1. If they are in E1, verify whether the hypotheses of Theorem 4.2.1 imply that CE1,/y admits finite limits (or that E1 ⊆ E2); if neither can be shown, Lemma 4.1.9 is invalid. An independent check: instantiate the proof in the minimal category with objects {x,y,z,z'}, E2-morphisms x→z, x→z', E1-morphisms z→y, z'→y, and no pullback z ×_y z'; test whether Kpt(τ) is still weakly contractible and whether the proof's limit step can be performed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central two-step proof of Theorem 1.0.3 depends on [1, Theorem 4.1.1] (here Theorem 3.3.1), whose weak-contractibility hypothesis is supplied for Theorem A by Proposition 4.1.8. Proposition 4.1.8 is not established: its proof via Lemma 4.1.9 requires finite limits in an overcategory of CE2, but the diagrams whose limits are taken have edges in E1. Indeed, Definition 4.1.7.4 declares morphisms of Kpt(τ) to be natural transformations that are pointwise in E1, so the horn maps ∂∆^m → Kpt(τ) induce diagrams in C/y whose 1-simplices lie in E1. Lemma 4.1.9 instead invokes Proposition B.0.2 for CE2,/y, which only accepts diagrams whose edges are in E2. No argument shows E1 ⊆ E2 or that CE1,/y admits finite limits. More broadly, the proof asserts 'CE2 admits pullbacks and CE2→C preserves pullbacks' without this following from the hypotheses of Theorem 4.2.1 (admissibility alone does not guarantee existence of pullbacks). The same missing pullback-completeness appears in Theorem 5.2.1, whose proof relies on Proposition 5.1.9 with the explicit hypothesis 'C admits pullbacks', absent from the theorem statement. Thus the load-bearing weak-contractibility step for pcomm is not proven, and the application of Theorem 3.3.1 is unjustified as written.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper aims to prove an ∞-categorical version of the Liu–Zheng gluing construction for exceptional pushforwards in abstract six-functor formalisms. The main theorem (Theorem 1.0.3) states that, under admissibility and a factorization condition on two edge collections E1, E2 in an ∞-category C, the natural map p : δ*_2 C^cart_{E1,E2} → C admits an extension along p for any target ∞-category D. The proof is split into two steps: Theorem 4.2.1 (extension along the map pcomm from commutative grids to C) and Theorem 5.2.1 (extension along the map pcart from Cartesian grids to commutative grids). The key technical input is the author's previous lifting theorem [1, Theorem 4.1.1] (restated as Theorem 3.3.1), whose weak contractibility hypothesis is to be supplied by Proposition 4.1.8 (for pcomm) and Proposition 5.1.9 (for pcart). The paper develops combinatorial tools: multisimplicial sets, the ∞-category of compactifications Kpt(τ), and the ∞-category of cartesianizations Kart(τ), together with inner anodyne inclusions proved in the appendices.","tokens_in":28440,"tokens_out":4002,"duration_ms":36267,"significance":"If the main theorem is correct, it reproduces in written detail a load-bearing gluing theorem that is currently only available in unpublished work of Liu–Zheng. The paper is valuable because it isolates a parameter-free combinatorial statement (Theorem 1.0.3) and separates the proof into two independent extension problems, each governed by a concrete simplicial set (compactifications and cartesianizations). It also contains substantial combinatorial material, such as the inner anodyne results in Appendix A and the limit existence criteria in Appendix B, that could be reused independently. The result would be a useful foundation for rigorous treatments of six-functor formalisms in derived algebraic geometry and p-adic geometry. However, the current manuscript has several unresolved technical gaps in exactly the weak-contractibility statements that carry the proof, so its significance is conditional on those gaps being repaired.","major_comments":[{"comment":"There is an inconsistency between the edge classes used to define morphisms of Kpt(τ) and the edge classes used in the cofilteredness proof. Definition 4.1.7.4 declares morphisms of Kpt(τ) to be natural transformations that are pointwise in E1. However, the proof of Lemma 4.1.9 takes limits in the overcategory C_{E2}/y, and Remark 4.1.10 explicitly constructs refinement maps in E2. Proposition B.0.2 only supplies finite limits in overcategories of C_{E2} when the diagrams involved have E2-edges, and it does not apply to diagrams whose edges are in E1 unless one knows E1 ⊆ E2 or that C_{E1}/y has finite limits. No such implication is established. Consequently, the proof of Lemma 4.1.9 does not prove that Kpt(τ) is cofiltered, and Proposition 4.1.8—which is the weak-contractibility hypothesis needed for Theorem 4.2.1—is not established as written.","section":"Lemma 4.1.9 and Definition 4.1.7.4"},{"comment":"The proof of Lemma 4.1.9 uses the assertion that 'CE2 admits pullbacks and CE2 → C preserves pullbacks' and that this follows from the hypotheses of the proposition. This does not follow from admissibility as defined in Definition 1.0.2: admissibility of E2 only says that E2 contains identities, is stable under pullbacks, and satisfies the right-cancellation property. It neither asserts that C admits pullbacks nor that the pullbacks of E2-edges exist. Indeed Theorem 4.2.1, which relies on Proposition 4.1.8, has no hypothesis that C admits pullbacks. Thus the proof of Lemma 4.1.9 invokes an unstated pullback-completeness assumption that is absent from the theorem statement.","section":"Lemma 4.1.9, proof"},{"comment":"Theorem 5.2.1 is stated without any assumption that C admits pullbacks, but its proof relies on Proposition 5.1.9, whose explicit hypothesis is 'If C admits pullbacks'. Proposition 5.1.9 is used to conclude that Kart(τ) is a contractible Kan complex, and this contractibility is then used in the proof of Theorem 5.2.1 to verify the weak-contractibility condition of Theorem 3.3.1. Since the hypothesis of Proposition 5.1.9 is not among the hypotheses of Theorem 5.2.1, the proof of Theorem 5.2.1 is incomplete as stated. The same missing assumption also affects the compatibility verification in the induction step over the truncation degree i, because the diagram in Eq. (106) only makes sense for squares that admit the required decompositions.","section":"Theorem 5.2.1, proof, and Proposition 5.1.9"},{"comment":"The proof of Theorem 5.2.1 proceeds 'by induction on i' and defines g'_cart as the union of the induced maps g^i_cart over i ≥ −2, but the argument does not specify the precise inductive statement nor verify that the extensions g^i_cart agree on the overlaps δ*_2 C^{i-1}_{E1,E2} ∩ δ*_2 C^{i}_{E1,E2} in a way that is compatible with the lifting problem. In particular, the diagram in Eq. (107) introduces a map p^{i-1,2}_{cart} : δ*_2 C^{-2}_{E1,E2} → δ*_2 C^{i-1}_{E1,E2} and then applies the induction hypothesis, but it is not shown that the extension produced for the (i−1)-level can be chosen so that the composed lifting problem with p^{i-1,i}_{cart} is solvable. This is a load-bearing point for the definition of g'_cart on the union, since the union in Eq. (80) is built from all truncation levels.","section":"Proof of Theorem B, induction on i"}],"minor_comments":[{"comment":"There are numerous typographical errors that impede reading, for example 'The abstract six-functor formalism plays is' in the first sentence of the introduction, 'vibration' instead of 'fibration' in the proof of Theorem 4.2.1, and 'thee case' in the road map. A careful proofreading pass is needed.","section":"Throughout"},{"comment":"The notation 'Fun_{E1,E2}(Cpt_n, C)' used in the proof of Proposition 4.1.8 is not defined before first use; the intended definition (functors sending horizontal arrows to E1 and vertical arrows to E2) should be stated explicitly at the point of definition.","section":"Definition 4.1.7"},{"comment":"In Proposition 5.1.15, the notation X appears in Eq. (108) as 'X =' followed by a blank; this is a missing definition or a typesetting error that makes the displayed formula unintelligible. Please correct and define X.","section":"Notation 5.1.13 and Proposition 5.1.15"},{"comment":"The paper cites [1, Theorem 4.1.1] as the main technical engine, but the numbering convention between the present paper and [1] is not synchronized; for instance, the road map discusses 'Theorem 4.2.1' and 'Theorem 5.2.1' before those theorems are stated, and at times refers to 'Section 3.3' when Theorem 3.3.1 is introduced. Please add cross-references and clarify the numbering.","section":"Section 3.2 and Theorem 3.3.1"},{"comment":"The definition of ǫ_n in Construction 5.1.19 uses the notation Λ^n_0 and µ^n_0 from Eq. (91), but the proof of Lemma 5.1.18 is only partially given; in particular, part (2) states the identities without proof. Please fill in this justification, since the pullback property in Claim 5.1.22 relies on these identities.","section":"Remark 5.1.20 and Construction 5.1.19"}],"recommendation":"major_revision","confidential_remarks":"The paper depends crucially on the author's own preprint [1, Theorem 4.1.1], which is not yet published and whose correctness could not be independently verified during this review. The editor may wish to ensure that [1] is available and, if possible, that an independent check of its main theorem has been obtained before the present paper is accepted. The main gaps identified in the major comments are, in my view, repairable either by adding hypotheses such as 'C admits pullbacks' or by adjusting the edge-class bookkeeping; they do not appear to be circular, but they currently block the central claim. For this reason I recommend major revision rather than rejection, with the expectation that the author provides a repaired proof of Proposition 4.1.8 and a corrected statement/proof of Theorem 5.2.1."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a serious written attempt at the Liu-Zheng gluing theorem, and the new combinatorial machinery is real work. But as printed, the central weak-contractibility proof does not go through: the edge-class bookkeeping in Proposition 4.1.8 is internally inconsistent, and the application of [1, Theorem 4.1.1] is not justified as written. An expert could probably repair it, but the paper needs revision before it can be cited in place of Liu-Zheng.\n\nWhat is genuinely new: the categories Kpt(τ) and Kart(τ), the inner anodyne inclusions □^n ⊂ Cpt^n and ⊞^n_cart ⊂ Cart^n, and the overall reduction of the gluing problem to the author's own lifting theorem. The proof is not circular: Theorems A and B are derived from [1, Theorem 4.1.1], and the target results are not assumed. The self-citation is heavy, but that alone is not a flaw.\n\nSoft spots, in order of seriousness:\n1. Proposition 4.1.8 and Lemma 4.1.9. Definition 4.1.7.4 says morphisms of Kpt(τ) are pointwise E1. The cofilteredness proof takes limits in C_{E2}/y of diagrams whose edges lie in E1. No argument shows E1 ⊆ E2 or that C_{E1}/y admits finite limits. This is exactly where the weak-contractibility hypothesis for [1, Theorem 4.1.1] is supposed to come from, so the hypothesis is unsupported as written.\n2. Theorem 5.2.1 applies Proposition 5.1.9, which explicitly assumes C admits pullbacks. That hypothesis is absent from Theorem 5.2.1 and from Theorem 1.0.3. Without pullbacks, the contractibility of Kart(τ) is not established.\n3. The compatibility condition with f′ is asserted in both proofs with “by construction” and “one checks,” but not actually checked. It may be checkable, but the text does not do it.\n4. There is no theorem-by-theorem comparison with Liu-Zheng [6,7], so it is hard to tell exactly which statements are being reproduced and where the simplifications occur.\n\nNone of these is obviously fatal; the architecture is plausible and the motivation is clear. But the paper currently does not deliver what it promises: a written proof that can be cited in place of Liu-Zheng. People working on six-functor formalisms who need the gluing theorem will want to read it, and the combinatorial core is worth preserving. It deserves a serious referee—send it out, but expect heavy revision.","headline":"A serious written attempt at Liu-Zheng's gluing theorem with genuinely useful combinatorial scaffolding, but the printed proof has a load-bearing gap in the weak-contractibility step and needs revision before it can serve as the citable replacement it aims to be.","tokens_in":29254,"tokens_out":2493,"would_cite":false,"duration_ms":24173,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["18N60","14F08"],"pacs":[],"model":"deepseek-v4-flash","headline":"Exceptional pushforwards exist in any admissible ∞-category","keywords":["six-functor formalism","exceptional pushforward","infinity-categories","gluing functors","compactifications","cartesian squares","admissible edges","lifting problems"],"falsifier":"Check Lemma 4.1.9 directly: for a 1-morphism with two compactifications, the construction takes a limit in the overcategory C_{E2}/y over a diagram whose edges lie in E1, then uses Remark 4.1.10 to produce refinement maps. If there is an admissible pair (E1,E2) for which those refinement maps land in E2 rather than E1, the cofilteredness proof breaks and weak contractibility of Kpt(τ) is no longer established, so Theorem A does not follow.","tokens_in":27903,"feed_emoji":"🔗","tokens_out":6182,"duration_ms":58212,"temperature":0.7,"pith_summary":"This paper proves an ∞-categorical gluing theorem: given two admissible families of edges E1 and E2 in an ∞-category C, any functor defined on grids whose squares are pullbacks extends along the diagonal projection to a functor on C itself. That extension is the abstract version of the exceptional pushforward construction, assembling f! from its values on open immersions and proper maps. The proof splits into two extension theorems, one for commutative grids and one for cartesian grids, and the combinatorial heart is showing that the relevant categories of compactifications and cartesianizations are weakly contractible. If correct, the paper supplies a fully written, simplified proof of a key unpublished gluing theorem used in abstract six-functor formalisms.","feed_headline":"Exceptional pushforwards exist in any admissible ∞-category","feed_subtitle":"A two-step lifting proof turns compactification grids into a written gluing theorem for six-functor formalisms.","key_machinery":"The load-bearing objects are two combinatorial ∞-categories attached to an n-simplex: Kpt(τ), whose objects are ways to realize an n-simplex as a grid with vertical edges in E1 and horizontal edges in E2, and Kart(τ), whose objects are right Kan extensions along an up-set map and which encode decompositions of commutative squares into pullback squares. The arguments show that both are weakly contractible and that the source inclusions □^n ⊂ Cpt^n and ⊞^n_cart ⊂ Cart^n are inner anodyne, so the lifting theorem from [1, Theorem 4.1.1] applies. This reduces the gluing problem to a purely combinatorial statement about partially ordered sets and their admissible edge classes.","core_discovery":"The central claim is Theorem 1.0.3: for an ∞-category C with admissible edge collections E1 and E2 such that every morphism factors as a morphism in E1 followed by one in E2 and every morphism in E1∩E2 is k-truncated, the lifting problem for the map p:δ∗2C^cart_{E1,E2}→C has a solution for every ∞-category D. The map p, induced by composing along the diagonal of an n×n grid, factors as pcart followed by pcomm, and the paper proves both extension theorems (Theorem 4.2.1 for pcomm and Theorem 5.2.1 for pcart) using a lifting criterion from the companion work. The load-bearing inputs are weak contractibility of the ∞-category Kpt(τ) of compactifications and of the cartesianization category Kart(τ), which encode all ways to decompose a simplex into E1-then-E2 directions and to decompose commutative squares into pullback squares. When both hold, any functor on the cartesian grids extends canonically, giving the exceptional pushforward in the abstract six-functor formalism.","pith_inferences":["If weak contractibility of Kpt(τ) holds, the same argument should extend to the symmetric setup with E1 and E2 interchanged, as long as the decomposition condition is adjusted; the paper does not state this explicitly.","The k-truncation hypothesis is likely removable from the pcomm half, since Theorem 4.2.1 never uses it; only the pcart step depends on it.","The proof strategy suggests that any six-functor formalism satisfying the two admissible-edge conditions automatically produces f! with coherent functoriality; verifying compatibility with composition of arbitrary morphisms in C would be a natural next step."],"forward_implications":["A functor defined on cartesian grids extends uniquely, up to contractible choice, to all of C, producing the exceptional pushforward in an abstract six-functor formalism.","The two-step factorization isolates the role of the k-truncated condition: it is used only in the pcart step, where it lowers the truncation level when moving from commutative to cartesian squares.","The result turns Nagata-style compactification decompositions into a formal property of ∞-categories with admissible edge classes, so it applies to any context admitting such classes, not just schemes.","The paper gives a written, simplified proof of an unpublished gluing theorem that previously circulated informally and is a core piece of current six-functor formalism constructions."],"supporting_citations":[{"why":"Supplies the lifting theorem (Theorem 4.1.1) used as the technical engine for both extension proofs.","marker":"[1]"},{"why":"Motivates the compactification formalism through the classical gluing of pseudofunctors.","marker":"[3]"},{"why":"Provides the original gluing construction and the notation that this paper adapts.","marker":"[6]"},{"why":"Sets the context of enhanced six operations for which this gluing statement is needed.","marker":"[7]"},{"why":"Furnishes the ∞-category foundations, inner anodyne maps, limits, and Kan extensions used throughout.","marker":"[8]"},{"why":"Supplies marked simplicial sets and the marked-categorical language for admissible edges.","marker":"[9]"}],"fun_headline_variants":["Compactification grids yield exceptional pushforward","Admissible edge collections define exceptional pushforward","Lifting compactifications to construct pushforward","Cartesian grids glue into exceptional pushforward","Weak contractibility implies exceptional pushforward"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything rests on Proposition 4.1.8, the claim that for every simplex the ∞-category Kpt(τ) of compactifications is weakly contractible; the written proof of that claim uses a cofilteredness argument in which the edge classes E1 and E2 are handled in a way that does not obviously match Definition 4.1.7.4.","fun_headline_variants_meta":{"raw":{"variants":["Compactification grids yield exceptional pushforward","Admissible edge collections define exceptional pushforward","Lifting compactifications to construct pushforward","Cartesian grids glue into exceptional pushforward","Weak contractibility implies exceptional pushforward"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000235,"raw_usage":{"total_tokens":1463,"prompt_tokens":868,"completion_tokens":595,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":531}},"tokens_in":484,"tokens_out":595,"duration_ms":6354,"temperature":1.0,"reasoning_tokens":531,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T22:42:27.754226+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check Lemma 4.1.9 directly: for a 1-morphism with two compactifications, the construction takes a limit in the overcategory C_{E2}/y over a diagram whose edges lie in E1, then uses Remark 4.1.10 to produce refinement maps. If there is an admissible pair (E1,E2) for which those refinement maps land in E2 rather than E1, the cofilteredness proof breaks and weak contractibility of Kpt(τ) is no longer established, so Theorem A does not follow.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Motivates the compactification formalism through the classical gluing of pseudofunctors."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies marked simplicial sets and the marked-categorical language for admissible edges."}],"review_version":1}