{"id":"f70b263f-4ad5-4ed1-a387-8323c2bfabfb","arxiv_id":"1908.09372","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For pseudo-coherent finite maps of schemes, this preprint proves the concrete dualizing pseudofunctor f^♭ is isomorphic to the abstract f^!, and gives explicit comparisons for tensor, Hom, and Koszul-regular immersions.","lead":"This paper works out, in detail, how two ways of defining Grothendieck duality for maps of schemes agree: an abstract category-theoretic construction and a concrete one built from homomorphisms and derived functors. The first installment covers finite pseudo-coherent maps and shows the resulting dualizing functors match, including their behavior under composition, base change, and tensor products.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The finite-map duality is plausible, but the pseudofunctoriality of (-)^♭ in §2.5 is asserted, not proved; that is the load-bearing gap, not Lemma 2.3.8.","rationale":"The reader identified Lemma 2.3.8 as the weakest assumption. I disagree: that lemma is standard and correctly cited; the construction would indeed break without it, but the risk is low. The real soft spot is the formal pseudofunctoriality of (-)^♭ in §2.5. The paper gives no proof of the cocycle condition for (2.5.1), and Proposition 2.5.2, which identifies (-)^♭ with the abstract (-)^×, rests entirely on that unproved formal verification. Because the central claim is a concrete realization of an abstract pseudofunctor, a failure here would directly invalidate the claimed equivalence, not just a peripheral compatibility. The same pattern appears in the Koszul-regular section, where the hardest pseudofunctoriality is outsourced to a companion preprint. These are addressable, not obviously wrong, so CONDITIONAL remains the right verdict. The proposed concrete check settles the §2.5 gap; if it passes, the finite-map core is solid. The reader's conditional recommendation is therefore unchanged.","tokens_in":81528,"tokens_out":17475,"duration_ms":155733,"concrete_test":"Check the pseudofunctoriality cocycle condition for (2.5.1): for a triple of pseudo-coherent finite maps W → X → Y → Z, write the pentagon comparing the two composite isomorphisms h^♭g^♭f^♭ ⇒ (fgh)^♭ obtained from (2.5.1) and verify it commutes using only the counits t_* and the standard isomorphism R(fgh)_* ≅ Rf_*Rg_*Rh_*. If the pentagon commutes, Proposition 2.5.2 is justified; if not, the concrete realization is not a pseudofunctor. As a secondary check, complete the details left to the reader in Lemma 3.5.3 and the sign computation in the final paragraph of §3.5, since Theorem 2.10.22 depends on this affine calculation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing concern is the unproved assertion in the first paragraph of §2.5 that the isomorphisms (2.5.1) make (-)^♭ a contravariant pseudofunctor on pseudo-coherent finite maps. The text says 'One verifies formally' and then Proposition 2.5.2 is a one-line 'Consequently' identifying (-)^♭ with the abstract (-)^×. No cocycle or pentagon condition for (2.5.1) is displayed. If the associativity constraint for (-)^♭ fails, then (-)^♭ is only an oplax functor and the objectwise adjunction of Corollary 2.3.6 does not automatically upgrade to an isomorphism of pseudofunctors. The central claim that the concrete construction is a realization of the abstract f^! would then be unjustified. Lemma 2.3.8, by contrast, is a standard cited boundedness result (Stacks tag 0A6H) and is not the weak point. A secondary related gap is that Theorem 2.10.22, needed for the Koszul-regular concrete realization, is reduced to [NkS19a, Appendix C.6] and several subdiagram verifications in §3.5 are left to the reader.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This first part of a planned two-part exposition compares the abstract twisted-inverse-image pseudofunctor f^! with a concrete construction f^♭ for pseudo-coherent finite scheme maps. For such f, the paper defines f^♭ := \\bar f^* RHom_{f_* O_X}(f_* O_X, -), proves a sheafified finite duality isomorphism RHom(Rf_* F, G) ≅ Rf_* RHom(F, f^♭ G) (Cor. 2.3.9), derives the resulting adjunction Rf_* ⊣ f^♭ (Cor. 2.3.10), and then develops tor-independent base change, tensor/Hom compatibility, a trace and fundamental-class formalism, and a Koszul-regular realization (Propositions 2.6.4, 2.7.7, 2.8.2, 2.9.7, 2.10.12, Theorem 2.10.22). The announced central claim is that f^♭ is a pseudofunctor and is isomorphic as a pseudofunctor to the abstract f^× over pseudo-coherent finite maps of qcqs schemes (Proposition 2.5.2). The final section translates the Koszul-regular case into commutative algebra and completes it via a computation over Koszul complexes (§3.5).","tokens_in":81733,"tokens_out":5702,"duration_ms":60902,"significance":"If the missing coherence verifications are supplied, this will be a valuable and highly detailed bridge between the concrete duality theory of Hartshorne's Residues and Duality and the abstract f^! theory, under weaker hypotheses than are usual in the literature. The paper's strengths include explicit diagram chases, clean reductions from global schemes to affine commutative algebra, a serious treatment of base-change and tensor/Hom compatibilities, and a nontrivial Koszul-regular computation in §3.5. I also agree with the external assessment that Lemma 2.3.8 is a standard, properly cited boundedness result (Stacks tag 0A6H) and is not the weak point of the paper. The weak point is instead that several load-bearing pseudofunctoriality and coherence claims are asserted rather than proved, so the central comparison theorem is not fully established as written.","major_comments":[{"comment":"The pseudofunctoriality of f^♭ is the hinge of the paper's central claim, but it is not proved. The text says 'One verifies formally' that the isomorphisms (2.5.1) make (-)^♭ a contravariant pseudofunctor, and Proposition 2.5.2 is then a one-line 'Consequently' identifying (-)^♭ with (-)^×. No associativity or unitality coherence diagrams for (2.5.1) are displayed, and no proof is given that the objectwise isomorphism ξ_f of Corollary 2.3.6 is compatible with composition of finite maps. If the coherence constraints for (2.5.1) were to fail, Corollary 2.3.9 would only give objectwise adjunctions and would not upgrade to an isomorphism of pseudofunctors. The later local discussion in §3.1.20–3.1.23 constructs π_{ξ,φ} and states that 'one checks' pseudofunctoriality, but again no cocycle condition for triple composites is verified. This is a load-bearing gap: please supply a complete proof of the pseudofunctor laws for (2.5.1), or give a precise reference that contains the full coherence proof.","section":"§2.5, especially (2.5.1) and Proposition 2.5.2"},{"comment":"The concrete Koszul-regular realization is claimed to be pseudofunctorial in Theorem 2.10.22, but the proof is not complete as written. Lemma 2.10.24 reduces the theorem to commutativity of subdiagrams A©–D©, but the proofs for B© and D© are deferred with phrases such as 'shown by arguments like those used above' and 'can be shown in the same way as the last diagram in [L09, 4.6.8]'. The final commutative-algebra proof of Proposition 3.5.1 also leaves many subdiagrams of (3.5.2) to 'obvious or straightforward' verification, and Lemma 3.5.3 says 'Details are left to the reader'. Since Theorem 2.10.22 is needed to identify the concrete ω_f ⊗ Lf^*(-) with f^♭ as a pseudofunctor, not merely objectwise, this is another load-bearing point. Please either complete the missing diagram verifications or state explicitly which parts are being imported from [NkS19a, Appendix C.6] with theorem/proposition numbers.","section":"§2.10, Theorem 2.10.22 and Lemma 2.10.24; §3.5, Proposition 3.5.1"},{"comment":"The paper repeatedly leaves substantial compatibility statements to the reader. Proposition 2.8.3 (transitivity of ζ) has a one-line 'Proof. Left to the reader.' The assertion in §2.3.12 is explicitly marked 'will not be used, so the proof is omitted.' Section 2.11 says 'Some details follow; the rest are left to the reader.' Some of these are indeed peripheral, but Proposition 2.8.3 concerns the pseudofunctoriality of a map that is advertised as a concrete interpretation of an abstract duality map, and the pattern of deferred verifications makes it impossible for a reader to check the central claims without reconstructing a substantial portion of the proof. At minimum, each deferred statement should be labeled as an exercise or a conjecture, and all statements used in the proof of Proposition 2.5.2 or Theorem 2.10.22 should be proved or explicitly referenced.","section":"§2.8.3, §2.3.12, §2.11"}],"minor_comments":[{"comment":"The sentence 'for any map f ∈ F, Theorem 2.3.10 holds' refers to a nonexistent 'Theorem 2.3.10'; the intended reference appears to be Corollary 2.3.10.","section":"§2.5, first paragraph"},{"comment":"The hypothesis 'RHom_X(Lf^*F, f^♭G) ∈ D_qc(Y)' appears to be a typo: the complex lives on X, so the intended condition should presumably be membership in D_qc(X).","section":"§2.9.11(iii)"},{"comment":"The paper uses a large number of 'unlabeled subdiagrams are clear/obvious/straightforward' reductions. Given the length and technical density of the diagram chases, it would greatly help the reader if each such reduction were accompanied by a one-line indication of which adjunction or naturality property is being used, or if the fully expanded diagrams were collected in an appendix.","section":"General presentation"}],"recommendation":"major_revision","confidential_remarks":"The manuscript leans heavily on the author's own [L09] and on the companion preprint [NkS19a], and Theorem 2.10.22 is reduced to [NkS19a, Appendix C.6] without a fully self-contained proof. For a journal submission, the editors may wish to require that the key pseudofunctoriality and coherence assertions be proved in the paper or that the external reference be published with precise statement and proof. The central idea is plausible and the expository apparatus is impressive, but the current text does not yet supply complete verification of its main comparison theorem."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short take: this is the real thing, but with an unproved formal verification at the weight-bearing point. Lipman removes noetherian hypotheses for the finite pseudo-coherent/perfect case, constructs f^♭ via RHom(f_*O_X, -), proves sheafified finite duality (Corollary 2.3.9), identifies f^♭ with the abstract f^× (Proposition 2.5.2), and gives a new proof of Koszul-regular pseudofunctoriality (Theorem 2.10.22). That is a genuine advance over Hartshorne and the author's own earlier notes, not just exposition.\n\nCredit where it is due: Section 2 has real content. The adjunction and duality statements come with detailed diagram chases and local reductions, and the affine translation in Section 3, including the explicit generator-level calculation for Koszul-regular immersions, is serious mathematical writing. The paper is also honest about what is deferred; it flags the reliance on [Co00] and on the companion preprint [NkS19a].\n\nSoft spots, in proportion to how soft they actually are. First, the first paragraph of §2.5 says 'One verifies formally' that (2.5.1) makes (-)^♭ a pseudofunctor, and Proposition 2.5.2 is then a one-line 'Consequently.' No associativity or cocycle diagrams are displayed. Since the paper's central claim is that the concrete construction realizes the abstract f^!, this is a real gap in presentation, even if not necessarily in substance. A referee should ask for that verification to be written out. Lemma 2.3.8 is not the weak point; it is standard and properly cited. Second, Theorem 2.10.22, the hardest compatibility statement, is reduced to [NkS19a, Appendix C.6] and to the material in §3.5, and the §3.5 proof leaves several subdiagram checks to the reader. That is acceptable but brittle, especially since the companion preprint is not yet published. Third, the abstract promises the full Ideal Theorem for essentially finite type maps, while this part only treats the finite pseudo-coherent case and says the rest is being prepared. The mismatch should be fixed in the abstract or otherwise made clear at the top. Fourth, there are several 'left to the reader' verifications throughout; individually minor, but they accumulate.\n\nBottom line: if the §2.5 pseudofunctoriality verification is supplied, this becomes a solid foundation piece. For people working in Grothendieck duality and derived categories of schemes, it is worth citing and worth refereeing. I would send it to a serious referee, and my own verdict would be conditional on completing that verification.","headline":"A careful, mostly convincing infrastructure paper that makes the finite pseudo-coherent case of the concrete f^♭ = abstract f^! dictionary precise, with one load-bearing pseudofunctoriality assertion that needs to be written out.","tokens_in":82296,"tokens_out":2027,"would_cite":true,"duration_ms":22435,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14F05"],"pacs":[],"model":"deepseek-v4-flash","headline":"For pseudo-coherent finite maps, the paper proves that the concretely defined functor f^♭ is a pseudofunctorial right adjoint to derived direct image and agrees with the abstract twisted inverse image.","keywords":["Grothendieck duality","twisted inverse image","pseudo-coherent finite maps","derived categories","fundamental class","Koszul-regular immersions","flat base change","sheaf-Hom duality"],"falsifier":"Find a pseudo-coherent finite map f: X → Y and a complex G in D^+_qc(Y) for which RHom_{O_Y}(f_*O_X,G) fails to be bounded below or fails to have quasi-coherent cohomology. Lemma 2.3.8 forbids exactly this, so such an example would destroy the D^+_qc-valued adjunction and the sheafified duality isomorphism built on it.","tokens_in":81288,"feed_emoji":"🧮","tokens_out":19030,"duration_ms":193506,"temperature":0.7,"pith_summary":"This first part of a larger project aims to prove that the abstract, category-theoretic twisted inverse image functor and the concrete, formula-driven version used in the classical theory are the same for finite scheme maps whose direct image is pseudo-coherent. Its main theorem constructs a functor f^♭ from the derived sheaf-Hom RHom_{O_Y}(f_*O_X,-) and proves a sheafified duality isomorphism Rf_*RHom_X(F,f^♭G) ≅ RHom_Y(Rf_*F,G), so that f^♭ is a right adjoint of Rf_* on bounded-below quasi-coherent complexes. It then shows this concrete right adjoint is pseudofunctorial and canonically isomorphic to the abstractly constructed f^×, so every structural compatibility—flat base change, tensor product, internal Hom, traces—can be studied by explicit formulas. A reader should care because this replaces a notoriously abstract theory with computable commutative algebra for a broad class of maps, dropping the noetherian hypotheses that the classical treatment needed. The paper's stated program is the same comparison for all essentially finite-type maps, and this part is the finite-map foundation for that program.","feed_headline":"Concrete twisted inverse image matches abstract duality","feed_subtitle":"For pseudo-coherent finite maps, a derived-Hom formula gives the right adjoint and its compatibilities.","key_machinery":"The load-bearing object is the concrete functor f^♭ defined through the factorization f = φ\\bar f, with \\bar f flat and φ a restriction of scalars. The key equivalence identifies D_qc(X) with D_qc(Y,f_*O_X) by way of R\\bar f_*, so a right adjoint for Rf_* is produced from the classical derived-Hom adjunction for modules: the right adjoint of restriction of scalars is RHom_{O_Y}(f_*O_X,-), equipped with its natural f_*O_X-module structure. A second load-bearing fact, Lemma 2.3.8, guarantees that this derived Hom stays in the bounded-below quasi-coherent derived category when fed a pseudo-coherent complex and a bounded-below quasi-coherent complex. The whole paper then runs on the resulting explicit adjunction and the bifunctorial duality map derived from it.","core_discovery":"The central claim, on the paper's own terms, is that the pseudo-coherent finite map case of Grothendieck duality has an explicit canonical form. For such an f, let \\bar f: X → (Y, f_*O_X) be the natural flat map and let φ be restriction of scalars along O_Y → f_*O_X. The paper sets f^♭G to be \\bar f^* of the module-theoretic right adjoint of φ_*, an object whose underlying O_Y-complex is RHom_{O_Y}(f_*O_X,G). The main sheafified theorem is the bifunctorial isomorphism Rf_*RHom_X(F,f^♭G) ≅ RHom_Y(Rf_*F,G) for all F in D_qc(X) and G in D^+_qc(Y), with D_qc in place of D^+_qc when f is perfect. From this it derives an adjunction Rf_* ⊣ f^♭, proves the adjunction is pseudofunctorial, and identifies f^♭ with the abstract f^× by a canonical isomorphism of pseudofunctors. The remainder of the paper gives concrete descriptions of the induced base-change, tensor, Hom, trace, and fundamental-class maps, including the Koszul-regular immersion formula ω_f = Hom_X(⋀^d f^*(I/$I^{2}$),O_X)[-d].","pith_inferences":["Extending beyond the paper, the same factorization-through-structure-sheaf strategy should produce explicit formulas for the smooth case in the announced continuation, with the smooth dualizing complex Ω^d_f[d] ⊗ f^*(·) playing the role that derived Hom plays here.","A testable extension is to drop quasi-compactness or separatedness whenever Lemma 2.3.8 still gives values in D^+_qc; nothing else in the construction appears to require them once the right adjoint is known to exist.","The proof that Koszul-regular pseudofunctoriality reduces to a check on one generator suggests that other intricate compatibility diagrams of duality may admit equally short commutative-algebra proofs.","Because f^♭ is built from derived Hom, the theory is immediately computable in affine examples, so one could test the advertised compatibility diagrams with concrete Koszul and matrix-factorization computations."],"forward_implications":["The abstract right adjoint f^× for a pseudo-coherent finite map is canonically f^♭, so the duality functor is no longer an existence statement but a formula.","The sheafified isomorphism Rf_*RHom_X(F,f^♭G) ≅ RHom_Y(Rf_*F,G) remains valid without noetherian hypotheses and with only bounded-below quasi-coherent complexes, extending the classical finite-map duality.","Tor-independent and flat base change for the concrete functor take an explicit form: the base-change map β_σ(G): v^*f^♭G → g^♭u^*G is an isomorphism in the independent-square situation, and it is realized concretely from resolutions.","For Koszul-regular closed immersions f, the functor f^♭G is canonically ω_f ⊗^L_X Lf^*G with ω_f = Hom_X(⋀^d f^*(I/I^2),O_X)[-d], so duality for such immersions is governed by the normal-bundle determinant.","For perfect affine maps, the standard trace map for perfect complexes supplies a fundamental class Lf^*G → f^♭G that is an isomorphism for finite étale maps, linking concrete duality to residues and traces."],"supporting_citations":[{"why":"gives the original finite-map construction and the Ideal Theorem program that the paper's concrete functor generalizes.","marker":"[H66]"},{"why":"supplies the abstract right-adjoint formalism, the projection and Hom compatibilities, and the diagram tools used to identify f^♭ with f^×.","marker":"[L09]"},{"why":"supplies the definition and theory of pseudo-coherent and perfect complexes, plus the trace map used for the fundamental class.","marker":"[Il71]"},{"why":"cited for the boundedness and quasi-coherence of RHom (Lemma 2.3.8) and for Koszul-regularity facts needed in section 2.10.","marker":"[St18]"},{"why":"used to characterize pseudo-coherent finite maps through pseudo-coherence of f_*O_X, the hypothesis of the main theorem.","marker":"[LN07]"},{"why":"an earlier proof of pseudofunctoriality for Koszul-regular immersions whose result the paper recovers by the concrete commutative-algebra reduction.","marker":"[Co00]"}],"fun_headline_variants":["Concrete twisted inverse image equals abstract for finite maps","Derived Hom formula gives explicit f^! for pseudo-coherent maps","RHom expression pins down f^! for pseudo-coherent finite maps","Grothendieck duality made explicit for pseudo-coherent finite morphisms","Pseudo-coherent finite maps: one formula unifies duality theories"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The construction rests on the cited lemma that the derived sheaf-Hom of a pseudo-coherent complex with a bounded-below quasi-coherent complex is again bounded-below and quasi-coherent; if that lemma were false, f^♭ would not take values in the category where the main duality isomorphism is stated.","fun_headline_variants_meta":{"raw":{"variants":["Concrete twisted inverse image equals abstract for finite maps","Derived Hom formula gives explicit f^! for pseudo-coherent maps","RHom expression pins down f^! for pseudo-coherent finite maps","Grothendieck duality made explicit for pseudo-coherent finite morphisms","Pseudo-coherent finite maps: one formula unifies duality theories"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001169,"raw_usage":{"total_tokens":4860,"prompt_tokens":993,"completion_tokens":3867,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":609,"completion_tokens_details":{"reasoning_tokens":3776}},"tokens_in":609,"tokens_out":3867,"duration_ms":25290,"temperature":1.0,"reasoning_tokens":3776,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:13:59.809371+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a pseudo-coherent finite map f: X → Y and a complex G in D^+_qc(Y) for which RHom_{O_Y}(f_*O_X,G) fails to be bounded below or fails to have quasi-coherent cohomology. Lemma 2.3.8 forbids exactly this, so such an example would destroy the D^+_qc-valued adjunction and the sheafified duality isomorphism built on it.","supporting_citations":[],"review_version":1}