{"id":"6457ed90-7c05-46da-9f41-65bd68455803","arxiv_id":"1908.01201","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A 2-categorical equivariant fundamental groupoid is defined, and its discrete quotient is shown to be Morita invariant for representable orbifolds.","lead":"The authors construct a 2-category version of the equivariant fundamental groupoid and prove it is invariant under Morita equivalence, so the discrete fundamental groupoid becomes a well-defined invariant of representable orbifolds. This gives orbifold topologists a new categorical invariant that does not depend on how the orbifold is presented as a group action.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Prop. 5.6's path-lifting steps rely on an unproved and generally false principal-bundle claim for X^H -> Xbar^Hbar, so one of the two generating cases for Theorem 5.1 is not established.","rationale":"The reader's weakest assumption was the cited generation theorem, Prop. 3.5 of [16]. I do not dispute that this is a necessary condition, but the more immediate blocker is internal: in the proof of Prop. 5.6, the map used to lift paths and homotopies is asserted to be the quotient of a free N-action on X^H. That assertion is false in general, and the finite example makes the failure explicit. The fact that the fixed set X^H may be empty for a subgroup H while its image in the quotient is Hbar-fixed shows the claimed equality p(X^H)=Xbar^Hbar is not a formality. If p fails to be a fibration on the nonempty fixed components, the equivalence of 2-categories for the first generating type is unproved, and Theorem 5.1 and Corollary 5.2 do not follow from the written proof. This reinforces the CONDITIONAL verdict rather than changing it; I would not reject the paper, since the theorem may be repairable with a correct slice-theoretic argument, but the current text does not supply it.","tokens_in":18395,"tokens_out":40510,"duration_ms":431324,"concrete_test":"Compute the two fixed sets in the finite example G=C2xC2, N=C2x{1}, K={1}xC2, H={(0,0),(1,1)}, X=G/K: X^H is empty but Xbar^Hbar is nonempty, so the asserted quotient/principal-fibration structure is absent. Then re-run Prop. 5.6's arrow-lifting step for a subgroup with nonempty X^H, for example H=K in the same model, and check whether lifting can be justified without a new slice-theoretic fibration argument. If no such argument exists, the proof of Theorem 5.1 remains conditional on that missing step.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 5, Prop. 5.6: the proof of the quotient-by-free-normal-subgroup generator contains an unsupported lifting claim. In the paragraphs 'Surjective on Arrows', 'Full on 2-cells', and 'Faithful on 2-cells', paths and homotopies in Xbar^Hbar are lifted to X^H because p:X^H -> Xbar^Hbar is said to be 'the quotient of the free action of N, so this is a principal fibration'. This is not justified: for x in X^H and n in N, nx is H-fixed only if n^{-1}Hn fixes x; N need not normalize H, so N does not generally act on X^H. The assertion is also false as a map of sets: with G=C2xC2, N=C2x{1}, K={1}xC2, H={(0,0),(1,1)}, and X=G/K, one has X^H=empty while Xbar^Hbar is a point, so p(X^H) is not all of Xbar^Hbar. Thus the fibration used to lift arrows and 2-cells is unavailable on the face of it. Since Prop. 5.6 is one of the two explicit types whose composition is asserted to prove Theorem 5.1, the Morita-invariance claim is not established by the written argument. This concern is independent of the cited generation theorem (Prop. 5.3); even if that theorem is accepted, the proof of the first generating case has a hole.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs a 2-categorical enhancement Π_G(X) of tom Dieck's equivariant fundamental groupoid for a compact Lie group G acting smoothly on a manifold X, together with a discrete quotient Π^d_G(X). The main theorem (Theorem 5.1) asserts that every equivariant essential equivalence of translation groupoids G⋉X → H⋉Y induces a weak equivalence Π_G(X) → Π_H(Y), and hence that Π^d_G(X) is a Morita invariant, i.e., an orbifold invariant for representable orbifolds. The proof reduces the statement, via a generation result quoted from the authors' prior work ([16, Prop. 3.5] here Prop. 5.3), to two generating forms of equivariant essential equivalence: quotient by a freely acting normal subgroup (Prop. 5.6) and extension from a subgroup to an induced action (Prop. 5.9).","tokens_in":18719,"tokens_out":31980,"duration_ms":309248,"significance":"If the result is correct, it gives a natural 2-categorical resolution of the non-invariance of tom Dieck's non-discrete equivariant fundamental category and yields a concrete invariant of representable orbifolds. The paper is clearly organized, carefully develops the two-dimensional Grothendieck construction, and includes instructive examples. The authors are also explicit about the representability hypothesis and about which parts of the argument are quoted from earlier work. However, the correctness of the central claim currently rests on two proof points that are not fully established: the principal-bundle lifting assertion in Prop. 5.6 and the completeness of the continuity argument in Lemma 5.8.","major_comments":[{"comment":"In the paragraphs 'Surjective on Arrows' and 'Full on 2-cells', the proof asserts that p : X^H -> Xbar^Hbar is 'the quotient of the free action of N, so this is a principal fibration' and uses this to lift paths and homotopies. This assertion is not justified and is in general false: N need not act on X^H, because for x in X^H and n in N one has h(nx)=nx for all h in H only if n^{-1}Hn fixes x, which is not implied by normality of N. A concrete instance is G=D_8, N=<r>, H=<s>, X=G/H; here N∩H=1 and N acts freely on X, but N does not preserve X^H={H,r^2H}, and p : X^H -> Xbar^Hbar is a 2-to-1 map onto a point rather than a principal N-bundle. Since the lifting of arrows and 2-cells in Prop. 5.6 is used to establish one of the two generating cases for Theorem 5.1, the Morita-invariance claim is not proved by the written argument.","section":"Section 5, Prop. 5.6"},{"comment":"The proof of Lemma 5.8 is an informal grid-pattern argument. The text asserts that locally defined adjustment elements can be patched together continuously, but the continuity at transition lines and at corners where several neighbourhoods meet is not actually verified, and the displayed formula for the diagonal extension is not fully specified. Since Lemma 5.8 is used in Prop. 5.9 to prove both fullness and faithfulness on 2-cells for the second generating case, this is load-bearing and needs a complete proof.","section":"Lemma 5.8"}],"minor_comments":[{"comment":"The passage says that identifying all arrows connected by 2-cells produces the category Π^d_G(X); a sentence explaining why this quotient is well-defined, i.e., that the equivalence relation is compatible with composition, would help the reader.","section":"Section 3, after Def. 3.3"},{"comment":"The sentence 'So the map (G/K)^H -> (G/K)^H is a principal fibration' appears to have a typo in the target; the target should be the corresponding quotient space (G/K)^Hbar.","section":"Section 5, Prop. 5.6, 'Full on 2-cells'"},{"comment":"The coherence conditions for the pseudo natural transformation are dismissed as 'a straightforward calculation'; since the rest of the proof is quite detailed, please at least state the associativity and unit constraints and indicate where the verification is carried out.","section":"Proposition 4.6"},{"comment":"In the presentation Z⊕Z/<abab^{-1},b^2>, the elements a and b are not defined; they should be defined explicitly.","section":"Example 5.12"},{"comment":"The notation '(e,(\\ell_1^{-1},\\zeta))' for the arrow in Π_L(X) is confusing; if the intended arrow is (\\ell_1^{-1},[\\zeta]) with target (L/K,y), it should be written clearly in the standard arrow notation.","section":"Prop. 5.9, 'Essentially Surjective on Arrows'"}],"recommendation":"major_revision","confidential_remarks":"The main obstacle is the unsupported principal-bundle claim in Prop. 5.6; if the authors can supply a correct lifting argument, the paper may well be salvageable. I also recommend asking them to make Lemma 5.8 fully rigorous, since the current sketch is hard to check. Finally, Theorem 5.1 depends heavily on Prop. 3.5 of [16]; I did not audit that theorem, but the authors should restate its hypotheses explicitly so that the reader can see exactly what is being assumed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know about this paper. First, the construction is genuinely nice: the authors build a 2-categorical version of tom Dieck's equivariant fundamental groupoid via a Grothendieck construction over the orbit 2-category, and the discrete quotient recovers tom Dieck's π^d_1. That part is well-motivated and cleanly presented. Second, the main invariance claim is not established as written: the proof of Proposition 5.6, the quotient-by-free-normal-subgroup generator, relies on lifting paths and homotopies along p: X^H → Xbar^Hbar, asserting this is a principal fibration for the free action of N. That is false. N need not act on X^H because it need not normalize H. Concretely, take G=C2×C2, N=C2×{1}, K={1}×C2, H the diagonal, and X=G/K. Then X^H is empty while Xbar^Hbar is a point, so p(X^H) is not all of Xbar^Hbar. The same issue recurs in 'Full on 2-cells' where the map (G/K)^H → (Gbar/Kbar)^Hbar is called a principal fibration. Since Prop 5.6 is one of the two generating cases for Theorem 5.1, the Morita invariance and the orbifold invariance of π^d_1 are not proved by the text.\n\nThere are smaller gaps: Proposition 4.6's coherence is deferred to 'a straightforward calculation,' and Lemma 5.8's grid-pattern argument is informal, with continuity asserted rather than shown. Those are minor by comparison.\n\nCredit where due: the 2-categorical setup is a real improvement, the functoriality in Section 4 is mostly solid, and the examples (5.10–5.12) illustrate the invariant well. The reliance on Prop 5.3 from the authors' earlier paper is not itself a problem; the problem is the unsupported lifting inside the first generator.\n\nI think the theorem may well be true—one can imagine fixing Prop 5.6 by assuming N normalizes H or by choosing different fixed-point data—but that fix is not on the page. A referee should be sent this, because the construction and the question are important and the gap is specific and repairable. I would not cite the invariance result until the proof is repaired.","headline":"The 2-categorical tom Dieck construction is clean and worth knowing, but the main invariance theorem is not proved as written because Prop 5.6's path-lifting claim is false.","tokens_in":19196,"tokens_out":4224,"would_cite":false,"duration_ms":40868,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["55Q91","18E15","57R18"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper constructs a 2-categorical equivariant fundamental groupoid and proves its discrete quotient is Morita invariant, giving an orbifold fundamental group for representable orbifolds.","keywords":["equivariant fundamental groupoid","orbifold invariant","Morita equivalence","translation groupoids","2-categories","Grothendieck construction","fundamental group"],"falsifier":"The claim would collapse if one found an equivariant essential equivalence φ: G⋉X → H⋉Y for which the induced functor on the discrete groupoids was not essentially surjective or not fully faithful, or found two translation groupoids representing the same orbifold whose discrete equivariant fundamental groupoids are not equivalent.","tokens_in":18212,"feed_emoji":"🔄","tokens_out":7191,"duration_ms":67705,"temperature":0.7,"pith_summary":"This paper targets a well-defined notion of fundamental group for orbifolds: spaces that are locally quotients of manifolds by group actions but may have points with nontrivial stabilizer. The authors take tom Dieck's equivariant fundamental groupoid for a Lie-group action and build a 2-category version, then prove that the discrete quotient of this 2-category is invariant under Morita equivalence, the standard equivalence relation that identifies different group-action presentations of the same orbifold. The payoff is that the discrete equivariant fundamental groupoid is an orbifold invariant for representable orbifolds, so topologists can use it to distinguish orbifolds the way the classical fundamental group distinguishes manifolds. If correct, the result gives a computable invariant that does not depend on which group action is used to present the orbifold.","feed_headline":"Discrete equivariant fundamental groupoid is an orbifold invariant","feed_subtitle":"A 2-category version of the fundamental groupoid stays well-defined under the equivalences that identify orbifolds.","key_machinery":"The load-bearing construction is the 2-categorical Grothendieck category ∫_{O_G} ΠX, where ΠX sends each orbit G/H to the fundamental groupoid of the fixed set X^H and each orbit map to the action-induced functor; this packages tom Dieck's equivariant fundamental groupoid as a 2-category whose 2-cells are homotopy classes of paths in orbit space. The proof then rests on a generation result: every equivariant essential equivalence of translation groupoids is a composite of two explicit forms, a quotient by a freely acting normal subgroup and an induced-action inclusion L⋉X → G⋉(G×_L X). For each form, path-lifting lemmas show that the induced functor is essentially surjective on objects, full on arrows, and full-and-faithful on 2-cells, which together give the weak equivalence.","core_discovery":"Defining Π_G(X) as the 2-dimensional Grothendieck construction of the fixed-set fundamental-groupoid functor over the orbit category, the paper shows that every equivariant essential equivalence φ: G⋉X → H⋉Y induces a weak equivalence Π(φ): Π_G(X) → Π_H(Y) of 2-categories. Quotienting each Π_G(X) by its 2-cells recovers exactly tom Dieck's discrete category π^d_1(G,X), so the quotient functor is a weak equivalence of categories. The paper concludes that π^d_1 is Morita invariant, and therefore a well-defined orbifold invariant for orbifolds representable as quotients of compact Lie group actions with finite isotropy.","pith_inferences":["If the conjecture that every orbifold is representable holds, the same argument would extend the invariance statement from representable orbifolds to all orbifolds, giving a fully general orbifold fundamental group.","Because Π_G(X) is a 2-category, its automorphism 2-groups or homotopy category could yield finer invariants that detect isotropy data which the discrete quotient π^d_1 collapses.","The path-lifting lemmas used for the two generating equivalences suggest a direct recipe for computing π^d_1 of a quotient orbifold from fixed-set data of a presentation, which may be implementable in concrete examples."],"forward_implications":["The discrete fundamental groupoid π^d_1(G,X) is independent of the choice of translation groupoid representing a representable orbifold, so it can be assigned to the orbifold itself.","The richer 2-category Π_G(X) carries more information than the discrete quotient and is itself Morita invariant, so it is a finer orbifold invariant.","Any zig-zag of equivariant essential equivalences, not just the two generating types, preserves the invariant, making the invariance stable under composing presentations.","For representable orbifolds, π^d_1 can serve as an obstruction: two orbifolds with inequivalent discrete equivariant fundamental groupoids are not Morita equivalent and hence not the same orbifold."],"supporting_citations":[{"why":"Defines the equivariant fundamental groupoid π_1(G,X) and the discrete version π^d_1 that this paper reformulates categorically and proves invariant.","marker":"[20]"},{"why":"Supplies Proposition 3.5, the generation result that every equivariant essential equivalence is a composite of the two explicit forms on which Theorem 5.1 depends.","marker":"[16]"},{"why":"Gives the 2-dimensional Grothendieck construction used to build the 2-category Π_G(X) from the fixed-set functor ΠX.","marker":"[5]"},{"why":"Establishes that smooth orbifolds can be represented by Lie groupoids up to essential equivalence, connecting Morita equivalence to orbifolds.","marker":"[14]"},{"why":"Provides the partial representability result for non-effective orbifolds that motivates the paper's restriction to representable orbifolds.","marker":"[9]"}],"fun_headline_variants":["2-category equivariant groupoid: an orbifold invariant","Morita invariant equivariant groupoid for orbifolds","Equivariant fundamental groupoid invariant for orbifolds","Orbifold invariant from 2-category equivariant groupoid","Discrete equivariant groupoid: orbifold invariant proven"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on the claim that every way of changing one group-action presentation of an orbifold to another can be built by repeating two moves: dividing out by a subgroup that acts without fixed points and passing to an action induced from a smaller group.","fun_headline_variants_meta":{"raw":{"variants":["2-category equivariant groupoid: an orbifold invariant","Morita invariant equivariant groupoid for orbifolds","Equivariant fundamental groupoid invariant for orbifolds","Orbifold invariant from 2-category equivariant groupoid","Discrete equivariant groupoid: orbifold invariant proven"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00045,"raw_usage":{"total_tokens":2156,"prompt_tokens":723,"completion_tokens":1433,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":339,"completion_tokens_details":{"reasoning_tokens":1346}},"tokens_in":339,"tokens_out":1433,"duration_ms":12417,"temperature":1.0,"reasoning_tokens":1346,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:20:05.672161+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"The claim would collapse if one found an equivariant essential equivalence φ: G⋉X → H⋉Y for which the induced functor on the discrete groupoids was not essentially surjective or not fully faithful, or found two translation groupoids representing the same orbifold whose discrete equivariant fundamental groupoids are not equivalent.","supporting_citations":[{"cited_title":"tom Dieck, Transformation Groups, Walter de Gruyter (1987)","cited_arxiv_id":null,"evidence_quote":"Defines the equivariant fundamental groupoid π_1(G,X) and the discrete version π^d_1 that this paper reformulates categorically and proves invariant."},{"cited_title":"Pronk and L","cited_arxiv_id":null,"evidence_quote":"Supplies Proposition 3.5, the generation result that every equivariant essential equivalence is a composite of the two explicit forms on which Theorem 5.1 depends."},{"cited_title":"Buckley, Fibred 2-categories and bicategories, Journal of Pure and Applied Algebra 218 (2014), pp","cited_arxiv_id":null,"evidence_quote":"Gives the 2-dimensional Grothendieck construction used to build the 2-category Π_G(X) from the fixed-set functor ΠX."},{"cited_title":"Moerdijk, D.A","cited_arxiv_id":null,"evidence_quote":"Establishes that smooth orbifolds can be represented by Lie groupoids up to essential equivalence, connecting Morita equivalence to orbifolds."},{"cited_title":"Henriques, D","cited_arxiv_id":null,"evidence_quote":"Provides the partial representability result for non-effective orbifolds that motivates the paper's restriction to representable orbifolds."}],"review_version":1}