{"id":"d18ec555-18a0-4855-a5cc-15b1fb07cc73","arxiv_id":"2411.12927","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The homeomorphism and mapping class groups of a stable surface have automatic continuity if and only if every end of the surface is telescoping.","lead":"For a large family of infinite-type surfaces called stable surfaces, this paper decides exactly when the group of homeomorphisms, and the related mapping class group, has the automatic continuity property: every algebraic homomorphism to a separable group is necessarily continuous. It also proves the analogous result for homeomorphism groups of stable Stone spaces, answering two open questions of Kathryn Mann.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The only-if direction of Theorem A leans on Theorem 4.31 imported from Domat [10]; the proof is sketched and the extension from compactly supported closures to arbitrary closed subgroups is the least secure step.","rationale":"I read the paper in good faith. The positive direction of Theorem A is largely self-contained: it gives explicit Steinhaus constants, handles the compact part via Edwards-Kirby fragmentation, and the topological lemmas about stable end spaces and tiles are detailed. The negative direction, however, is a dependency chain ending at Theorem 4.31, which is imported from the second author's earlier paper and proved only by a sketch in Section 4.8. The reader's weakest assumption identifies exactly this imported theorem and the stabilizer hypotheses. I agree with that assessment. The concern is not an internal inconsistency; it is a correctness risk in an external black box. If Theorem 4.31 fails in the needed generality, the classification collapses to the positive direction. The stabilizer constructions in Section 4.9 are plausible, and the enlargement argument preserves nondisplaceability, but the terse Case (ii) and the reliance on [10] justify a conditional verdict. A focused re-derivation of Theorem 4.31 for the specific stabilizer subgroups would settle the issue; if it succeeds, the paper's central claim would be fully supported, and the verdict could be upgraded to ACCEPT.","tokens_in":25773,"tokens_out":16838,"duration_ms":182319,"concrete_test":"Take the special case of Theorem 4.31 needed for Theorem A: G = Stab(γ) for a separating curve γ isolating a non-telescoping end x, with the sequence {K_i} constructed in the three cases of Lemma 4.32. Independently re-derive the proof from [10, Sections 7, 8, 10] for this G, verifying: (1) the BBF projection axioms for the collection {g(K_i) : g in Stab(γ)} using [10, Lemma 3.8], including the fact that the K_i themselves are disjoint and only their translates overlap; (2) that the quasimorphism counting the sequence (i!) is defined on all of Stab(γ) and has unbounded value on f_{K,A}; (3) that f_{K,A} generates a copy of Q in H^1(Stab(γ);Z) with finite subproducts in the kernel. If any of these steps uses the full group generated by compactly supported mapping classes, Theorem 4.31 is not available in the needed generality.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem A's negative direction is a chain: Lemma 4.32 produces a non-telescoping end, Section 4.9 constructs a Stab(γ)-nondisplaceable sequence, and Theorem 4.31 converts that sequence into a discontinuous homomorphism to Q. The last link is the load-bearing one. The proof of Theorem 4.31 says that [10, Theorem 7.1], stated for closures of compactly supported mapping classes, works for any closed G with a G-nondisplaceable sequence because the projection axioms in [10, Lemma 3.8] only use pairwise overlapping subsurfaces. But the Ki in a nondisplaceable sequence are pairwise disjoint; the relevant overlaps are between G-translates of the Ki. One must check that the resulting Bestvina-Bromberg-Fujiwara projection complex is hyperbolic and that the quasimorphism counting the exponents ai is well-defined on all of G, not just on the compactly supported closure. The H^1(G;Z) contains Q step and the claim that finite subproducts map trivially are likewise imported from [10, Section 8]. The paper flags that this is a sketch, so the gap is acknowledged rather than hidden, but if the projection axioms or homology argument fail for the stabilizer subgroups of Section 4.9, the classification reduces to the positive direction. The stabilizer checks that the constructed curves or pairs of pants enlarge to a G-nondisplaceable sequence are themselves brief; Case (ii) in particular ('points in E′(y) play the role of genus') is terse, though the enlargement argument is plausible.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper gives a complete classification, for stable orientable surfaces without boundary, of when Homeo(Σ) and Map(Σ) have automatic continuity: they do exactly when every end is telescoping (isolated puncture, Cantor type, or a non-isolated genus-accumulating successor with Cantor-type predecessors). The positive direction is proved through a five-step fragmentation/commutator/pigeonhole argument with explicit Steinhaus constants (96 for Stone spaces, 288 for surface neighborhoods, 4896 for the final surface theorem), using an Eilenberg–Mazur swindle for surface bricks. The negative direction constructs discontinuous homomorphisms to Q by importing a theorem of Domat on nondisplaceable sequences in mapping class groups, after showing that a non-telescoping end leads to a stabilizer subgroup with such a sequence. The paper also proves automatic continuity for homeomorphism groups of stable second-countable Stone spaces and discusses an unstable example where the question remains open.","tokens_in":26059,"tokens_out":5250,"duration_ms":55572,"significance":"If Theorem A is correct, it resolves two questions of Mann in the stable case and provides the first complete classification for a large class of infinite-type surfaces. The framework developed in Sections 3 and 4 is modular and likely reusable: the fragmentation, commutator, and pigeonhole steps are organized as a general five-step program, and the Stone-space theorem (Theorem B) is a standalone result with the corollary on countable ordinals. The positive direction is a genuine technical advance, and the paper is explicitly honest about which parts are imported: the negative direction relies on a theorem of Domat [10] whose proof is only sketched. The explicit Steinhaus constants are a useful feature. The main risk is the extension of Domat's theorem from the closure of compactly supported mapping classes to arbitrary closed subgroups, which is load-bearing for the 'only if' direction.","major_comments":[{"comment":"The proof asserts that [10, Theorem 7.1], stated for the closure of compactly supported mapping classes, holds for any closed subgroup G with a G-nondisplaceable sequence. The provided justification is a sketch: it says the projection axioms are verified in [10, Lemma 3.8] for 'pairwise overlapping finite-type subsurfaces', but the subsurfaces Ki in Definition 4.30 are pairwise disjoint. The relevant overlaps are between G-translates of the Ki, and one must show that the resulting Bestvina–Bromberg–Fujiwara projection complexes are hyperbolic and that the quasimorphism coarsely counting the exponents ai is well-defined on all of G, not just on the closure of compactly supported classes. The same gap affects the later claims that fK,A generates a copy of Q in H1(G;Z) and that finite subproducts are trivial in homology. Since this theorem is the mechanism producing the discontinuous homomorphism to Q, the 'only if' direction of Theorem A is not fully established without a complete proof or a precise citation covering arbitrary closed subgroups.","section":"Section 4.8, Theorem 4.31"},{"comment":"After proving that the pairs of pants Pn are Stab(γ)-nondisplaceable, the proof says that one can 'expand these pairs of pants to subsurfaces of sufficiently high complexity and pass to a disjoint subsequence to obtain a Stab(γ)-nondisplaceable sequence.' No construction or proof is given that the expanded subsurfaces remain Stab(γ)-nondisplaceable, are pairwise disjoint, and are all homeomorphic to a fixed finite-type surface K of sufficient complexity. The argument for Pn uses the fact that a translate cannot cross certain boundary curves; enlarging Pn can create room for a translate to intersect the enlarged part while avoiding the original pair of pants. This step is load-bearing for the failure direction whenever Lemma 4.32 Case (iii) occurs.","section":"Section 4.9, Case (iii)"},{"comment":"The conclusion that x 'has finitely many predecessors and hence is a successor' is asserted without proof. The finiteness of the union of maximal equivalence classes in the annuli does not immediately imply that every predecessor of x is comparable to one of finitely many incomparable predecessors, as required by Definition 4.5. The comparability and maximality of the classes appearing in the annuli need to be argued explicitly, because this dichotomy is what routes the proof into Case (ii) of Theorem A.","section":"Lemma 4.32, Case 1"}],"minor_comments":[{"comment":"There is a typo: 'Fruedenthal compactification' should be 'Freudenthal compactification'.","section":"Theorem 4.29, Step 1"},{"comment":"The word 'uncontable' should be 'uncountable'.","section":"Section 4.6, Lemma 4.24"},{"comment":"There are several typos: 'f E(yi)' should likely be 'If E(yi)', and 'E(y0)' should be 'E(yi)' in the sentence about accumulation points.","section":"Lemma 4.15, proof"},{"comment":"The phrase '2 2ℵ0 many discontinuous homomorphisms' is missing formatting; it should read '2^{2^{ℵ0}}'.","section":"Section 4.8, Theorem 4.31"},{"comment":"The multi-brick/multi-neighborhood extensions are stated with only 'mutatis mutandis' or by reference to the single-brick proofs. This is acceptable, but the surface version in Proposition 4.28 involves boundaries and marked annuli, so a sentence indicating which steps change would improve readability.","section":"Proposition 3.17 and Proposition 4.28"},{"comment":"In item 2, the phrase 'the surface with a infinite genus' should be 'the surface with infinite genus'.","section":"Introduction, Example 1.3"}],"recommendation":"major_revision","confidential_remarks":"The positive direction is detailed and seems sound; the main uncertainty is the negative direction, which depends on an extension of Domat's theorem that is only sketched. If the authors can supply a complete proof or a precise reference covering arbitrary closed subgroups, the result would be a strong addition. Given that one of the authors is also an author of [10], this should be feasible. The editor may wish to have the BBF/projection-complex portion checked by an expert in the relevant literature."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bestvina–Domat–Rafi prove that for a stable, orientable boundaryless surface, Homeo(Σ) and Map(Σ) have automatic continuity iff every end is telescoping, and that Homeo(X) has automatic continuity for every stable Stone space. If the classification is right, it closes two questions of Mann and gives the first complete AC classification for a large class of infinite-type surfaces.\n\nThe positive half is the real meat and it is good. The Five Step Program (fragmentation, commutators via the Anderson trick and an Eilenberg–Mazur swindle, diagonalization, pigeonhole, wrap-up) is a genuinely reusable framework. The proof is largely self-contained and the explicit Steinhaus constants (96, 288, 4896) make the argument checkable. The use of the swindle to handle the absence of global shift maps on surfaces is clever and new, at least to me. The Stone space theorem is a clean byproduct, and Corollary 1.4 on countable ordinals is a nice concrete payoff.\n\nThe soft spot is exactly where the reader's report puts it: the only-if direction. Theorem 4.31 is imported from Domat [10] and the proof here is a sketch. The extension from closures of compactly supported mapping classes to arbitrary closed subgroups is stated in one sentence, and the stress-test concern is fair: you need the BBF projection complex to be hyperbolic and the quasimorphism counting the exponents to be well-defined on all of G, not just on the compactly supported closure. That is not a trivial step. The stabilizer nondisplaceability checks in Section 4.9 are also brief; Case (ii) ('points in E′(y) play the role of genus') is genuinely terse. I don't see a contradiction, and the authors flag the sketch honestly, but I would not call the negative direction fully verified from this paper alone. A referee should check Domat's argument and its applicability to the specific stabilizers.\n\nMinor: Proposition 3.17's 'mutatis mutandis' for multibricks is a bit quick, but it's not load-bearing. The reliance on Mann–Rafi [16] and Domat [10] is real but published and not circular.\n\nWho should read this: anyone working in geometric group theory or descriptive set theory with an interest in automatic continuity or big mapping class groups. It deserves a serious referee. I would send it to review and ask for a careful verification of the Domat import and the stabilizer sequence checks.","headline":"The positive direction is solid and the framework is reusable; the negative direction hinges on a sketched import from Domat [10] that needs checking before I'd call the classification fully verified.","tokens_in":26606,"tokens_out":2844,"would_cite":true,"duration_ms":28976,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57K20","57S05"],"pacs":[],"model":"deepseek-v4-flash","headline":"For stable surfaces, automatic continuity of the homeomorphism and mapping class groups holds exactly when every end is telescoping.","keywords":["automatic continuity","mapping class groups","infinite-type surfaces","homeomorphism groups","stable surfaces","telescoping ends","Steinhaus property","Stone spaces"],"falsifier":"Test the stabilizer built in case (iii) of Lemma 4.32: compute whether the sequence of pairs of pants satisfies the overlap assumptions of Theorem 4.31, and check whether the infinite product of a pseudo-Anosov map with exponents (i!) lies in the commutator subgroup of the stabilizer. If either fails, the discontinuous homomorphism to Q is not produced and the only-if direction collapses. Equally decisive would be finding one stable surface with a non-telescoping end whose mapping class group nevertheless has automatic continuity.","tokens_in":25541,"feed_emoji":"🗺️","tokens_out":7710,"duration_ms":78980,"temperature":0.7,"pith_summary":"The paper gives a complete classification of automatic continuity for homeomorphism and mapping class groups of stable surfaces: the group has the automatic continuity property precisely when every end of the surface is telescoping. Telescoping means each end is an isolated puncture, of Cantor type, or accumulates genus and is a successor whose predecessors are all of Cantor type. The positive direction is proved through a general five-step framework for the Steinhaus property, which fragments homeomorphisms into brick-supported pieces, writes pieces as commutators, and then reassembles them. The negative direction builds discontinuous homomorphisms to Q from sequences of non-displaceable finite-type subsurfaces. The same framework also establishes automatic continuity for homeomorphism groups of stable second countable Stone spaces, giving the first complete answers, under stability, to two open questions in the area.","feed_headline":"Homeomorphism groups of stable surfaces: continuity iff all ends telescope","feed_subtitle":"Telescoping ends—punctures, Cantor ends, and genus-accumulated successors—are exactly the good case.","key_machinery":"The central machinery is the five-step Steinhaus framework built around the notion of a brick: a countably infinite union of pairwise disjoint big annuli (in the surface case) or clopen pieces (in the Stone-space case) whose complement contains all relevant types of ends and whose homeomorphism type is independent of which infinite subcollection is chosen. The five steps are fragmentation of a supported map into brick-supported maps, realization of brick-supported maps as finite products of commutators via an infinite-product swindle, diagonal selection of a good sub-brick, a pigeonhole argument moving any brick into that good brick, and a final wrapping-up argument using Baire category and compactness. On the negative side, the load-bearing mechanism is a G-nondisplaceable sequence: pairwise disjoint finite-type subsurfaces, each homeomorphic to a fixed surface that supports a pseudo-Anosov map, each intersecting every image of itself under a closed subgroup G of the mapping class group. Projection-complex techniques turn such a sequence into a quasimorphism coarsely counting powers, yielding an element whose homology class generates a copy of Q and therefore a discontinuous homomorphism to Q.","core_discovery":"The central claim is Theorem A: if Σ is a connected, stable, orientable surface without boundary, then Homeo(Σ) and Map(Σ) have automatic continuity if and only if every end of Σ is telescoping, meaning each end is an isolated puncture, is of Cantor type, or is not isolated in the space of ends accumulated by genus and is a successor with all predecessors of Cantor type. The paper also claims Theorem B: if X is a stable second countable Stone space, then Homeo(X) has automatic continuity. Instead of checking homomorphisms directly, the proofs verify the Steinhaus property and invoke the standard consequence that Steinhaus implies automatic continuity. In the positive direction, a telescoping end admits a decomposition of a neighborhood into homeomorphic big annuli, called bricks, and a shift-and-swindle mechanism writes any supported homeomorphism as a bounded product of elements from a prescribed dense set. In the negative direction, a non-telescoping end produces a stabilizer containing a non-displaceable sequence, and an imported theorem from the literature turns such a sequence into a discontinuous homomorphism to Q, which then extends to discontinuous homomorphisms from Map(Σ) and Homeo(Σ).","pith_inferences":["A natural extension the authors leave implicit is to test the five-step recipe on other Polish groups of homeomorphisms admitting locally homogeneous telescoping neighborhoods and enough shift maps; the recipe predicts Steinhaus for any such group.","All known obstructions in this paper have Q inside the target group. If the literature conjecture that torsion-free separable groups without Q are always safe codomains is correct, non-displaceable sequences would be the only possible source of discontinuity, strengthening the classification philosophy.","The unstable surface constructed in Section 4.10, whose ends are all of Cantor type but whose maximal end types form countably many incomparable Cantor sets, is a sharp test case: deciding automatic continuity there would show whether stability can be relaxed to a weaker local condition.","For colored Stone spaces with infinitely many colors, stability fails and automatic continuity is unknown; a classification of color-preserving homeomorphism groups would be a natural Stone-space analogue of Theorem A."],"forward_implications":["Automatic continuity for stable surfaces is now completely decided: all telescoping ends give Steinhaus, and every non-telescoping end gives a discontinuous homomorphism to Q.","The groups Homeo(Σ) and Map(Σ) admit unique Polish group topologies whenever Σ is stable and all its ends telescope.","The same framework proves automatic continuity for homeomorphism groups of all stable second countable Stone spaces, including all countable ordinal spaces.","The discontinuous homomorphisms in the negative direction factor through the mapping class group, so the failure of automatic continuity is visible at the level of mapping classes rather than only through homeomorphism dynamics."],"supporting_citations":[{"why":"Supplies the Steinhaus-property criterion that every automatic-continuity proof in the paper ultimately invokes.","marker":"[24]"},{"why":"Establishes the Steinhaus strategy for homeomorphism groups of compact manifolds that the five-step program extends.","marker":"[14]"},{"why":"Provides the big-mapping-class-group results and neighborhood-decomposition techniques for surfaces with Cantor and finite sets removed that the positive direction builds on.","marker":"[15]"},{"why":"Defines stability and the partial order on ends of surfaces used throughout the classification.","marker":"[16]"},{"why":"Is the source of Theorem 4.31, the non-displaceable-sequence construction of discontinuous homomorphisms to Q used for the negative direction.","marker":"[10]"},{"why":"Supplies projection complexes whose actions verify the quasimorphism and commutator-length control behind Theorem 4.31.","marker":"[4]"},{"why":"Provides the fragmentation result used to decompose maps close to the identity on compact pieces.","marker":"[11]"},{"why":"Gives the infinite-product swindle used to write brick-supported surface homeomorphisms as finite products of commutators.","marker":"[3, 19]"}],"fun_headline_variants":["Stable surfaces: continuity iff all ends telescope","Automatic continuity classified for stable surface homeos","Telescoping ends decide homeomorphism group continuity","Stable surfaces mapped: continuity exactly when ends telescope","New framework yields continuity for stable Stone spaces too"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole negative half rests on an imported theorem stating that a closed subgroup of the mapping class group containing a sequence of disjoint finite-type surface pieces, each forced to overlap every image of itself under the subgroup, must admit a discontinuous homomorphism to Q; only a sketch of that theorem is given here, so if its geometric overlap assumptions fail for the particular stabilizers built from a non-telescoping end, only the positive half of the classification stands.","fun_headline_variants_meta":{"raw":{"variants":["Stable surfaces: continuity iff all ends telescope","Automatic continuity classified for stable surface homeos","Telescoping ends decide homeomorphism group continuity","Stable surfaces mapped: continuity exactly when ends telescope","New framework yields continuity for stable Stone spaces too"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000166,"raw_usage":{"total_tokens":1223,"prompt_tokens":887,"completion_tokens":336,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":503,"completion_tokens_details":{"reasoning_tokens":264}},"tokens_in":503,"tokens_out":336,"duration_ms":3909,"temperature":1.0,"reasoning_tokens":264,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T17:03:13.987880+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Test the stabilizer built in case (iii) of Lemma 4.32: compute whether the sequence of pairs of pants satisfies the overlap assumptions of Theorem 4.31, and check whether the infinite product of a pseudo-Anosov map with exponents (i!) lies in the commutator subgroup of the stabilizer. If either fails, the discontinuous homomorphism to Q is not produced and the only-if direction collapses. Equally decisive would be finding one stable surface with a non-telescoping end whose mapping class group nevertheless has automatic continuity.","supporting_citations":[{"cited_title":"Automatic continuity of homomorphisms and fixed points on metric compacta","cited_arxiv_id":null,"evidence_quote":"Supplies the Steinhaus-property criterion that every automatic-continuity proof in the paper ultimately invokes."},{"cited_title":"Automatic continuity for homeomorphism groups and applications","cited_arxiv_id":null,"evidence_quote":"Establishes the Steinhaus strategy for homeomorphism groups of compact manifolds that the five-step program extends."},{"cited_title":"Automatic continuity for homeomorphism Groups and big mapping class groups","cited_arxiv_id":null,"evidence_quote":"Provides the big-mapping-class-group results and neighborhood-decomposition techniques for surfaces with Cantor and finite sets removed that the positive direction builds on."},{"cited_title":"Large-scale geometry of big mapping class groups","cited_arxiv_id":null,"evidence_quote":"Defines stability and the partial order on ends of surfaces used throughout the classification."},{"cited_title":"Big pure mapping class groups are never perfect","cited_arxiv_id":null,"evidence_quote":"Is the source of Theorem 4.31, the non-displaceable-sequence construction of discontinuous homomorphisms to Q used for the negative direction."},{"cited_title":"Constructing group actions on quasi-trees and applications to mapping class groups","cited_arxiv_id":null,"evidence_quote":"Supplies projection complexes whose actions verify the quasimorphism and commutator-length control behind Theorem 4.31."},{"cited_title":"Edwards and Robion C","cited_arxiv_id":null,"evidence_quote":"Provides the fragmentation result used to decompose maps close to the identity on compact pieces."}],"review_version":1}