{"id":"2bd6e778-f3e7-4d87-a2ca-0c955b881347","arxiv_id":"2412.14398","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors construct the first examples of irreducible closed 4-manifolds admitting exotic diffeomorphisms, using a families Seiberg-Witten constraint and explicit lattice automorphisms.","lead":"This paper proves that certain irreducible 4-manifolds, spaces that cannot be split into simpler pieces, admit exotic diffeomorphisms: invertible maps that are topologically deformable to the identity but not smoothly. It solves a long-open question in 4-manifold topology and produces many new non-trivial boundary Dehn twists.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the argument is a sound reduction to the published families Seiberg-Witten constraint, with no internal gap that would threaten the main theorems.","rationale":"I read the paper as doing exactly what it claims: it reduces the existence of exotic diffeomorphisms on irreducible 4-manifolds to a previously established families Seiberg-Witten constraint, and then supplies the needed geometric input through realizable lattice automorphisms and a boundary Dehn twist commutator. The reader's weakest-assumption analysis correctly identifies the external theorem as the crux. My own pass looked for places where the written proof might be internally inconsistent. The only nontrivial point I found is the implicit identification of the spin^c structure s in Theorem 5.2 with one preserved by the f_i from Proposition 3.3. This is not stated, but it follows because the canonical class is w*t with w nonzero, so any lattice automorphism preserving the canonical class preserves t; since all basic classes in Eq. (2) are multiples of t, the chosen s is preserved. The OCR of the arXiv text makes the sign in the congruence in Theorem 5.2 look ambiguous, with a possible plus sign where the derivation gives a minus, but the published PDF should resolve this; I do not treat the OCR artifact as a mathematical objection. Overall I found no load-bearing concern, so I recommend leaving the reader's ACCEPT verdict unchanged.","tokens_in":13431,"tokens_out":41694,"duration_ms":374586,"concrete_test":"Verify in the published text of [4, Corollary 1.3] that the compact base B is allowed to have H^1(B;Z)!=0, in particular B=T^2, and confirm that all hypotheses hold for the T^2-family E constructed in Theorem 2.3 from the assumed isotopy of the commutator; if these checks pass, the contradiction establishing exoticity is valid.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The single load-bearing input is Theorem 2.2, quoted from the authors' published paper [4, Corollary 1.3]: for a compact family of spin^c 4-manifolds with odd Seiberg-Witten invariant, b+=3 mod 4 and b1=0, one has c1(D_E)=w2(H+(E)) mod 2. Both Theorem 1.2 and Theorem 1.4 derive a contradiction from this identity, so a hidden restriction there would collapse the results. I have not found such a restriction. The proof of Proposition 2.1 is consistent: the Serre spectral sequence degenerates, the divisibility by 32 is used correctly, and the conclusion c1(D_E)=0 mod 2 follows. The lattice construction in Proposition 3.3 correctly yields commuting automorphisms that preserve the canonical class and the orientation of H+, and the computation of w2(H+)=x1x2 is valid. The boundary Dehn twist construction is also internally sound, and Lemma 4.1's computation correctly forces c1(D_E)!=0 mod 2. One implicit point is that the spin^c structure s chosen in Theorem 5.2 need not be the canonical one, whereas Proposition 3.3 is stated for s0; this is harmless because the canonical class is a nonzero multiple of the primitive fiber class t, so any isometry preserving the canonical class fixes t and hence fixes every basic class r*t. Thus no adjustment to the reader's verdict is needed.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves that certain irreducible 4-manifolds admit exotic diffeomorphisms, providing the first known examples. Specifically, for logarithmic transforms E(4m)_{i,j} of elliptic surfaces with (i,j) outside an explicit finite set, and for complete intersections with c1 and σ divisible by 32, the natural map π0(Diff^+(X)) → Γ(X) does not split, so a commutator of two diffeomorphisms is exotic. The proof combines a families Seiberg-Witten constraint from the authors' prior work [4, Cor. 1.3] with a computation showing c1(D_E) = 0 mod 2 for families over a surface when σ and c1(s) are divisible by 32, and a construction of diffeomorphisms with specified cohomology actions via realization theorems of Lönne and Ebeling–Okonek. The same families constraint is used to detect non-trivial boundary Dehn twists on spin 4-manifolds with S^3 boundary, giving many new examples.","tokens_in":13632,"tokens_out":27181,"duration_ms":206504,"significance":"The main result resolves a long-standing open question: irreducible 4-manifolds can admit exotic diffeomorphisms. The method is a clean reduction to a published families invariant and avoids the usual dimensional obstruction to families invariants for irreducible manifolds. The boundary Dehn twist results are also new and substantially extend the known examples. The arguments in Sections 2–4 are carefully written and the theoretical framework is sound, with explicit constructions for the diffeomorphisms. If the example-verification issues in Section 5 are fixed, this will be an important contribution to 4-manifold topology.","major_comments":[{"comment":"The proof of Lemma 5.1 is incomplete. After reducing to j ≤ 15 and handling (1,15), (1,13), (1,11), the proof states that the remaining cases are (5,9), (7,9), (3,7), (5,7). However, many other pairs with j ≤ 15 that are not in S1 or S2, such as (3,11), (3,13), (5,11), (5,13), (7,11), (7,13), (9,11), (11,13), (7,15), and (11,15), are not addressed. Since this lemma is used in Theorems 5.2 and 5.3 to find spin^c structures with the required divisibility and odd Seiberg-Witten invariant, the proof of the elliptic surface examples is not complete as written. The authors should supply a complete finite verification (e.g., a systematic case check or computer-assisted verification) or a general argument covering all pairs.","section":"Lemma 5.1"},{"comment":"The congruence used to find the spin^c structure has a sign error. From (4mij - 4ijk0 - 2ja - 2ib - i - j) ≡ 0 mod 32, dividing by 2 gives ja + ib + 2ijk0 ≡ 2mij - (i+j)/2 mod 16, not 2mij + (i+j)/2. The same error appears in Theorem 5.3, where the correct constant is (2m-1)ij - (i+j)/2. This does not invalidate the theorems, since Lemma 5.1 is stated for every integer c, but the displayed equations in the proofs should be corrected.","section":"Theorem 5.2 and Theorem 5.3"}],"minor_comments":[{"comment":"The assertion that L0 contains at least three copies of H is not justified. It follows because L0 is an even indefinite unimodular form with b+(L0) = b+(X)-1 ≥ 3, so its standard decomposition contains at least b+(L0) hyperbolic planes; a brief explanation would help.","section":"Proposition 3.3"},{"comment":"The spin^c structure s chosen in these theorems need not be the canonical structure s0 used in Proposition 3.3. The application is harmless because any isometry preserving c1(s0) fixes the primitive fiber class t and hence preserves every basic class r·t, including c1(s); this point should be stated explicitly.","section":"Theorems 5.2 and 5.3"},{"comment":"The derivation of p_g ≡ 3 mod 4 is compressed. It follows from c1(X)^2 - σ(X) = 8χ(X) and the divisibility assumptions, which give 8χ(X) ≡ 0 mod 32 and hence χ(X) ≡ 0 mod 4; spelling this out would improve readability.","section":"Theorem 5.4"},{"comment":"Typo: 'the result will follow form Theorem 5.4' should read 'from Theorem 5.4'.","section":"Corollary 5.7"}],"recommendation":"major_revision","confidential_remarks":"The central machinery of the paper appears sound and the main theorems are likely correct. The main issues are in Section 5: the incomplete finite verification in Lemma 5.1 and the sign error in the congruence in Theorems 5.2 and 5.3. Both are fixable, but the lemma's incomplete case analysis is load-bearing and must be addressed before the paper can be accepted. The reliance on [4, Corollary 1.3] as a black box is acceptable since it is published, but the authors may want to state clearly that the examples depend on that theorem."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Baraglia and Konno settle a question that has been open since Ruberman's work: irreducible 4-manifolds can indeed admit exotic diffeomorphisms. The main theorems give explicit families—logarithmic transforms of elliptic surfaces and certain complete intersections—where the map from π0(Diff+) to Γ(X) does not split, with the exotic diffeomorphism realized as a commutator of two diffeomorphisms realizing commuting lattice automorphisms. They also give many new non-trivial boundary Dehn twists.\n\nThe proof is a clean reduction. The key input is Theorem 2.2, the authors' earlier families Seiberg–Witten constraint: for a family of spin^c 4-manifolds with odd SW invariant, b+=3 mod 4, b1=0, one has c1(D_E)=w2(H+(E)) mod 2. The paper proves Proposition 2.1, a divisibility computation showing c1(D_E)=0 mod 2 when σ(X) and c1(s) are divisible by 32, and then constructs commuting diffeomorphisms whose induced bundle H+ has non-zero w2. The contradiction is sharp. I checked the Serre spectral sequence step and the lattice realization: they are sound. The extension to the boundary Dehn twist is also careful; the explicit commutator in the 4-ball that equals the boundary Dehn twist is a nice piece of elementary topology, and Lemma 4.1 correctly uses the normal bundle to force c1(D_E) non-zero mod 2.\n\nSoft spots are minor. The paper leans heavily on Theorem 2.2 from the authors' own earlier work; the reader's weakest assumption is exactly that theorem, but the stress test found no hidden hypothesis, and the theorem is published and parameter-free, so this is acceptable. Lemma 5.1's finite checks are left as 'can be checked directly,' which is fine for a handful of pairs but slightly annoying. The complete intersection examples rely on standard facts about SW invariants of general type; those are cited. None of this threatens the main theorems.\n\nWho is this for? Anyone working in 4-manifold topology or mapping class groups. It deserves a serious referee; the referee's time would be well spent checking the application of Theorem 2.2 to the specific families and the finite checks. I would cite this paper.","headline":"Settles the long-open question of exotic diffeomorphisms on irreducible 4-manifolds with a clean reduction to the authors' published families Seiberg–Witten constraint.","tokens_in":14251,"tokens_out":1475,"would_cite":true,"duration_ms":12784,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57K40","57R50","57R57"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that certain irreducible 4-manifolds — logarithmic transforms of elliptic surfaces and complete intersections with c1 and σ divisible by 32 — admit exotic diffeomorphisms: diffeomorphisms topologically isotopic to the…","keywords":["exotic diffeomorphisms","irreducible 4-manifolds","families Seiberg-Witten invariants","mapping class group","boundary Dehn twist","elliptic surfaces","complete intersections","spin 4-manifolds"],"falsifier":"Try to exhibit a smooth isotopy from the commutator [f1,f2] constructed in Section 3 to the identity on any of the listed manifolds. The paper proves this is impossible: a smooth isotopy would assemble the mapping tori of f1 and f2 into a smooth $T^{2}$-family to which Proposition 2.1 forces c_1(\\mathcal{D}_E)=0 mod 2, while Theorem 2.2 together with the construction forces c_1(\\mathcal{D}_E)=w_2(H^+(E))≠0 mod 2. Concretely, one could attempt to compute the two sides of the identity c_1(\\mathcal{D}_E)=w_2(H^+(E)) mod 2 for such a family and look for a mismatch, which would indicate that the assumed isotopy cannot exist.","tokens_in":13151,"feed_emoji":"","tokens_out":11665,"duration_ms":88277,"temperature":0.7,"pith_summary":"The paper proves that irreducible closed smooth 4-manifolds can admit exotic diffeomorphisms: diffeomorphisms that are topologically isotopic to the identity but not smoothly isotopic. This resolves an open problem, because all previously known exotic diffeomorphisms of 4-manifolds lived on manifolds that split as connected sums. The examples are specific logarithmic transforms of elliptic surfaces, E(4m)_{i,j} with i and j odd and coprime outside a small exceptional set, and complete intersections whose first Chern class and signature are divisible by 32. The argument shows that the natural homomorphism from the group of orientation-preserving diffeomorphisms to the automorphism group of the intersection form does not split, and the exotic diffeomorphism is a commutator [f1,f2] of two diffeomorphisms with prescribed actions on cohomology. The same machinery yields many new examples of non-trivial boundary Dehn twists on spin 4-manifolds with $S^{3}$ boundary.","feed_headline":"Irreducible 4-manifolds can admit exotic diffeomorphisms","feed_subtitle":"Diffeomorphisms that are topologically trivial but not smoothly trivial are constructed on minimal complex surfaces.","key_machinery":"The load-bearing mechanism is a constraint from families Seiberg-Witten theory, quoted from the authors' earlier work: for any smooth family E→B of spin^c 4-manifolds with fiber X satisfying b_+(X)=3 mod 4 and b_1(X)=0, and for a spin^c structure with odd Seiberg-Witten invariant, one has c_1(\\mathcal{D}_E)=w_2(H^+(E)) mod 2, where \\mathcal{D}_E is the families index of the spin^c Dirac operator and H^+(E) is the maximal positive-definite subbundle of the cohomology bundle. The countervailing fact, proved in the paper, is that whenever σ(X) and c_1(s) are divisible by 32, any family over a closed orientable surface has c_1(\\mathcal{D}_E)=0 mod 2. The contradiction is produced by choosing two commuting diffeomorphisms f_1,f_2 whose induced automorphisms of $H^{2}$(X;Z) make the associated $T^{2}$-family have w_2(H^+)≠0. A lattice-theoretic lemma places c_1(s) inside a hyperbolic summand of the intersection form, and explicit sign-change automorphisms (diagonal −1 entries) generate the required w_2; known realization theorems lift these automorphisms to diffeomorphisms of elliptic surfaces and complete intersections.","core_discovery":"On the paper's own terms, the main theorem asserts that if X is either a logarithmic transform E(4m)_{i,j} with m≥1, j≥i≥1, i and j odd and coprime with (i,j)∉S1, or a complete intersection with c1(X) and σ(X) divisible by 32, then X admits an exotic diffeomorphism. This is established through a stronger structural statement: the surjective homomorphism π0(Diff^+(X))→Γ(X) does not split, meaning some automorphism of the intersection form in the image cannot be lifted to a diffeomorphism in a way compatible with the group structure. The exotic diffeomorphism is exhibited as the commutator [f1,f2] of two orientation-preserving diffeomorphisms whose cohomology actions commute, and the proof runs by building a smooth $T^{2}$-family from a hypothetical isotopy of that commutator and deriving a contradiction between two families Seiberg-Witten identities. The paper also proves a parallel theorem for boundary Dehn twists: for any compact simply-connected spin 4-manifold with b+=3 mod 4, an odd Seiberg-Witten invariant, c1(s) divisible by 32, and σ(X)=16 mod 32, the Dehn twist on X with an open ball removed is non-trivial in the relative mapping class group.","pith_inferences":["The construction is a template: any pair of commuting orientation-preserving diffeomorphisms of a spin 4-manifold with b_+=3 mod 4 and odd Seiberg-Witten invariant, whose cohomology actions produce a bundle with w_2(H^+)≠0, should yield an exotic commutator; the divisibility-by-32 conditions are the paper's device for making c_1(\\mathcal{D}_E) vanish and may be stronger than necessary.","The same T^2-family obstruction could detect non-splitting in the smooth mapping class group beyond commutators, for instance by using families over higher-genus surfaces where more commuting diffeomorphisms are available.","The boundary Dehn twist results suggest that the relative smooth mapping class group of a spin 4-manifold with S^3 boundary frequently has a Z_2 factor generated by the Dehn twist, and that this may interact with pseudo-isotopy and topological vs. smooth isotopy questions in dimension four."],"forward_implications":["The mapping class group of each listed manifold does not split over its image in Aut(Q_X), so there are cohomology automorphisms that cannot be realized by diffeomorphisms in a group-compatible way.","Exotic diffeomorphisms occur in both minimal elliptic surfaces and surfaces of general type, so the phenomenon is not an accident of one geography of complex surfaces.","For the listed spin 4-manifolds with S^3 boundary, the boundary Dehn twist is non-trivial in the relative mapping class group, providing relative exotic diffeomorphisms.","Because each exotic diffeomorphism is a commutator, it acts trivially on cohomology, so the phenomenon is invisible to the intersection form and requires families Seiberg-Witten theory to detect."],"supporting_citations":[{"why":"Supplies the central obstruction identity c1(D_E)=w2(H^+(E)) mod 2 for families with odd Seiberg-Witten invariant; every main result reduces to it.","marker":"[4]"},{"why":"Provides the proposition that the T^2-family assembled from f1 and f2 admits a families spin^c structure restricting to s on the fibers.","marker":"[1]"},{"why":"Realizes the required cohomology automorphisms as orientation-preserving diffeomorphisms of minimal elliptic surfaces.","marker":"[19]"},{"why":"Realizes the analogous automorphisms as diffeomorphisms of complete intersections.","marker":"[6]"},{"why":"Gives the topological isotopy result used to show the commutator is topologically isotopic to the identity once it acts trivially on cohomology.","marker":"[25]"},{"why":"Identifies the kernel of the relative-to-absolute diffeomorphism group map as generated by the boundary Dehn twist, setting up the boundary results.","marker":"[9]"},{"why":"Supplies the known context that the boundary Dehn twist is smoothly trivial for non-spin manifolds and always topologically trivial.","marker":"[23]"},{"why":"Gives the Seiberg-Witten invariants of logarithmic transforms of elliptic surfaces used to select spin^c structures with odd invariant and c1 divisible by 32.","marker":"[7]"},{"why":"Supplies the odd Seiberg-Witten invariant of the canonical spin^c structure on complete intersections of general type.","marker":"[21]"},{"why":"Supplies the classification of elements in indefinite unimodular even lattices used in Lemma 3.1 to move c1(s) into a hyperbolic summand.","marker":"[27]"}],"fun_headline_variants":["Exotic diffeomorphisms found on irreducible 4-manifolds","First exotic diffeomorphisms on irreducible 4-manifolds","Minimal complex surfaces admit exotic diffeomorphisms","Commutator trick yields exotic diffeomorphisms","Exotic diffeomorphisms via commutators on 4-manifolds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole argument rests on Theorem 2.2, taken from the authors' previous paper, which asserts that every smooth family of simply-connected 4-manifolds with b_+=3 mod 4 and odd Seiberg-Witten invariant satisfies c_1(\\mathcal{D}_E)=w_2(H^+(E)) mod 2; if that theorem carries a hidden hypothesis, or fails to apply to the specific $T^{2}$ families built from a hypothetical isotopy, both the exoticity results and the boundary Dehn twist results collapse.","fun_headline_variants_meta":{"raw":{"variants":["Exotic diffeomorphisms found on irreducible 4-manifolds","First exotic diffeomorphisms on irreducible 4-manifolds","Minimal complex surfaces admit exotic diffeomorphisms","Commutator trick yields exotic diffeomorphisms","Exotic diffeomorphisms via commutators on 4-manifolds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001267,"raw_usage":{"total_tokens":5160,"prompt_tokens":891,"completion_tokens":4269,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":507,"completion_tokens_details":{"reasoning_tokens":4181}},"tokens_in":507,"tokens_out":4269,"duration_ms":27175,"temperature":1.0,"reasoning_tokens":4181,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T12:17:45.012613+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Try to exhibit a smooth isotopy from the commutator [f1,f2] constructed in Section 3 to the identity on any of the listed manifolds. The paper proves this is impossible: a smooth isotopy would assemble the mapping tori of f1 and f2 into a smooth $T^{2}$-family to which Proposition 2.1 forces c_1(\\mathcal{D}_E)=0 mod 2, while Theorem 2.2 together with the construction forces c_1(\\mathcal{D}_E)=w_2(H^+(E))≠0 mod 2. Concretely, one could attempt to compute the two sides of the identity c_1(\\mathcal{D}_E)=w_2(H^+(E)) mod 2 for such a family and look for a mismatch, which would indicate that the assumed isotopy cannot exist.","supporting_citations":[{"cited_title":"On the Bauer-Furuta and Seiberg-Witten invari- ants of families of 4-manifolds","cited_arxiv_id":null,"evidence_quote":"Supplies the central obstruction identity c1(D_E)=w2(H^+(E)) mod 2 for families with odd Seiberg-Witten invariant; every main result reduces to it."},{"cited_title":"Obstructions to smooth group actions on 4-manifolds from families Seiberg-Witten theory","cited_arxiv_id":null,"evidence_quote":"Provides the proposition that the T^2-family assembled from f1 and f2 admits a families spin^c structure restricting to s on the fibers."},{"cited_title":"On the diffeomorphism groups of elliptic surfaces","cited_arxiv_id":null,"evidence_quote":"Realizes the required cohomology automorphisms as orientation-preserving diffeomorphisms of minimal elliptic surfaces."},{"cited_title":"On the diffeomorphism groups of certain algebraic surfaces","cited_arxiv_id":null,"evidence_quote":"Realizes the analogous automorphisms as diffeomorphisms of complete intersections."},{"cited_title":"Isotopy of 4-manifolds","cited_arxiv_id":null,"evidence_quote":"Gives the topological isotopy result used to show the commutator is topologically isotopic to the identity once it acts trivially on cohomology."},{"cited_title":"The stable mapping class group of simply connected 4- manifolds","cited_arxiv_id":null,"evidence_quote":"Identifies the kernel of the relative-to-absolute diffeomorphism group map as generated by the boundary Dehn twist, setting up the boundary results."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the Seiberg-Witten invariants of logarithmic transforms of elliptic surfaces used to select spin^c structures with odd invariant and c1 divisible by 32."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the odd Seiberg-Witten invariant of the canonical spin^c structure on complete intersections of general type."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the classification of elements in indefinite unimodular even lattices used in Lemma 3.1 to move c1(s) into a hyperbolic summand."}],"review_version":1}