{"id":"883de007-0619-4fbe-bea5-06689e40f1b7","arxiv_id":"2508.21057","paper_version":1,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A degree-one proper map between non-compact surfaces has a geometric kernel or is properly homotopic to a homeomorphism, if it is injective on Brown's proper fundamental group at every end; a new cπ1 invariant yields further sufficient conditions.","lead":"This paper proves new conditions under which a proper map between non-compact surfaces must send a non-contractible simple loop to a null-homotopic loop, called a geometric kernel. It introduces an end-based invariant, cπ1, that packages free homotopy classes of loops near an end, and applies it to degree-one maps and planar surfaces.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 3 as stated is false: S'=R^2 is not excluded, and a proper degree-one map R^2→R^2\\{0} satisfies all stated hypotheses but has no geometric kernel and is not properly homotopic to a homeomorphism.","rationale":"The reader correctly identified the end-wise monomorphism condition as delicate and noted internal contradictions in Example 3.3, but the single most load-bearing issue is that the paper's advertised Theorem 3 is false as stated. The body's Theorem 4.15 is protected by the standing assumption that neither surface is R^2, but the Introduction and Abstract do not carry that assumption for the domain. A concrete, elementary counterexample—the degree-one proper map from the plane to the punctured plane that is the identity outside the unit disk and constant inside—satisfies every stated hypothesis of Theorem 3 while violating the conclusion. This is not a subtle proof gap but a direct falsification of the central claim. The paper could be repaired by adding S′≠R^2 to Theorem 3 and the Abstract, and by fixing the contradictory Example 3.3; however, as currently written, the main advertised theorem is not correct.","tokens_in":42561,"tokens_out":22255,"duration_ms":239527,"concrete_test":"Verify the counterexample: define f:R^2→R^2\\{0} by f(x)=x for |x|≥1 and f(x)=0 for |x|<1. Check (1) f is proper, (2) deg(f)=1 by evaluating at a regular value such as (2,0), (3) π0(f) is the identity on one-point end spaces, (4) the end-wise monomorphism condition holds because π1(R^2,a)=0, (5) no non-contractible simple loop exists in R^2, and (6) the proper fundamental groups of R^2 and R^2\\{0} are 0 and Z, respectively, preventing proper homotopy to a homeomorphism. If all six checks pass, Theorem 3 as stated is false and must be amended to include S′ not homeomorphic to R^2.","verdict_should_be":"REJECT","load_bearing_attack":"The abstract and Introduction state Theorem 3 with only 'S is not homeomorphic to R^2', omitting the assumption that S′ is also not R^2. The Section 4 setup for Theorem 4.15 explicitly assumes neither M nor N is R^2, so the body proves a weaker statement than the advertised central claim. The omitted case is not harmless: let f:R^2→R^2\\{0} be defined by f(x)=x for |x|≥1 and f(x)=0 for |x|<1. This is a proper degree-one map: the preimage of any compact set is compact, and a regular value outside the unit disk has exactly one preimage with local degree 1. It induces the identity on the one-point end spaces, so π0(f) is injective. For the unique end of R^2, the Brown proper fundamental group π1(R^2,a) is trivial, so the end-wise monomorphism condition holds vacuously. Yet R^2 contains no non-contractible simple loop, so f has no geometric kernel, and R^2 is not homeomorphic to R^2\\{0}, so f is not properly homotopic to a homeomorphism. Thus Theorem 3 as stated in the abstract/Introduction is false. The reader's weakest_assumption concerned the end-wise π1 condition; the real load-bearing problem is a missing hypothesis in the central claim.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies geometric kernels of proper maps between non-compact orientable surfaces. A map has a geometric kernel if it sends a non-contractible simple loop to a null-homotopic loop. The author uses Brown's proper fundamental group to prove a cut-off theorem (Theorem 4.12): under end-wise monomorphism hypotheses, a proper map is, outside a compact set, a finite-sheeted covering. This is applied to degree-one maps to prove Theorem 4.15, stated in the Introduction as Theorem 3: under degree-one, end-injectivity, and an end-wise proper-π1 monomorphism condition, a geometric kernel exists unless the map is properly homotopic to a homeomorphism. The paper also introduces cπ1, a set of free homotopy classes of sequences of loops near an end, proves it coincides with conjugacy classes in Brown's proper fundamental group (Theorem 5.4), and uses it to give conditions for pinching handles (Theorems 5.5, 5.11) and a dichotomy for planar surfaces (Theorem 5.9).","tokens_in":42952,"tokens_out":26261,"duration_ms":303560,"significance":"If the main theorems are correct, the paper would give a useful extension of Edmonds' and Gabai's compact-surface geometric-kernel results to the non-compact setting, and the cπ1 framework is a natural invariant for end phenomena. The paper contains substantial original proof constructions, including a detailed cut-off argument and a 3-manifold analogue, and it explicitly builds on classical results of Nielsen, Edmonds, Brown-Tucker, and Epstein. However, the central advertised claim as stated in the abstract and Introduction is false, and an internal contradiction in Example 3.3 undermines the necessity discussion. Once the statement of Theorem 3 is corrected to match the body's standing assumptions and the example is repaired, the paper would be a meaningful contribution; in its current form the advertised result is not what is proved.","major_comments":[{"comment":"Theorem 3 as stated in the abstract and Introduction is not the theorem proved in the body. The standing assumptions of Section 4 explicitly require that neither M nor N is homeomorphic to R^2 ('Let M and N be connected, non-compact 2-manifolds ... neither M nor N is homeomorphic to R^2'), and Theorem 4.15 is proved under that setup. The Introduction states only S ≠ R^2. The omitted S' = R^2 case is a genuine counterexample: identify 0 and ∞ in S^2, let q: S^2 → S^2/{0∼∞} be the quotient, choose a homeomorphism h of S^2 fixing ∞ with h(0) not in {0,∞}, and let f = q∘h restricted to S^2∖{∞} ≅ R^2. This is a proper degree-one map R^2 → R^2∖{0}. Since π1(R^2,a)=0, the end-wise Brown-π1 monomorphism condition is vacuous, and π0(f) is injective. But R^2 has no non-contractible simple loop and is not homeomorphic to R^2∖{0}, so the conclusion of Theorem 3 fails. The body's Theorem 4.15 must be","section":"Abstract / Introduction (Theorem 3) / Section 4 preamble"},{"comment":"Example 3.3 is internally inconsistent. It opens by asserting 'such that ker π1(f) ≠ 0' and ends with 'On the other hand, π1(f) is injective'. These statements are mutually exclusive, since ker π1(f)=0 is exactly injectivity of π1(f). The example is used to demonstrate the necessity of π0-injectivity in Theorem 3, so this contradiction is load-bearing. Either the initial claim should be changed to π1(f) injective (with the example reframed accordingly) or the final injectivity assertion is false and the proof of it is wrong. As written, the example cannot support the claimed necessity discussion.","section":"Example 3.3"}],"minor_comments":[{"comment":"The notation for the ordinary fundamental group and Brown's proper fundamental group is not visually distinguished in the typeset text; in the paragraph after Theorem 1 the two symbols appear identical, making the sentence 'if π1(f) is a monomorphism (resp. isomorphism), then π1(f) is as well' nonsensical. Please use a distinct symbol (e.g., an underlined π1) for the proper fundamental group consistently.","section":"Throughout"},{"comment":"In the paragraph after defining the extension f: S' → S, the text says 'We claim that q has no geometric kernel', but q was originally only the map P → A; the claim should refer to the extended map f.","section":"Example 3.1"},{"comment":"The sentence containing 'or −[η′] = [η′] = [γ′_r] + [γ′_s]' appears to contain a typo; the intended relation should be stated clearly.","section":"Proof of Theorem 5.11"}],"recommendation":"major_revision","confidential_remarks":"The false advertised statement of Theorem 3 is a serious issue: the abstract and Introduction claim a theorem that is false as stated, even though the body proves a corrected version. The contradiction in Example 3.3 adds further doubt to the examples section. These are fixable within the manuscript's scope by adding S'≠R^2 to the statement and repairing the example, so I do not recommend rejection solely on these grounds. However, the current text cannot be accepted without revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — worth a read, but read the body, not the abstract. What is genuinely new: the cπ1(S,e) invariant (germ/free homotopy classes of sequences of loops near an end), its conjugacy-class characterization in Theorem 5.4, the bordered-surface classification in Theorem 4.8, the cutoff theorem 4.12, and the planar trichotomy 5.9. If those proofs hold, this is the first systematic non-compact extension of the simple-loop-conjecture program and deserves referee time.\n\nThe soft spots are real. The abstract and Introduction state Theorem 3 with only S not R2 excluded; the Section 4 setup for Theorem 4.15 assumes neither M nor N is R2. So the advertised central claim is stronger than what is proved. The counterexample I was shown does not rescue the criticism: the map R2→R2\\{0} sending the unit disk to 0 is not valued in R2\\{0}, and a one-ended plane cannot have injective π0 into the two-ended punctured plane. So I am not claiming Theorem 3 is false; I am claiming it is unproved as stated, and the paper should either add S′≠R2 or show the omitted case is vacuous.\n\nMore troubling: Example 3.3 and Example 3.4 each assert ker π1(f) ≠ 0 and then finish by saying π1(f) is injective. That is a flat contradiction. Presumably one of the two statements is a slip, but as submitted the examples cannot play their motivating role. There is also a garbled sentence in the Introduction comparing π1(f) with itself. These are fixable, but they lower confidence.\n\nI did not independently verify all the transversality and Nielsen–Edmonds steps; the strategy is coherent, and the citations to Brown–Tucker and Edmonds look appropriate. The self-citation [7] is used as background and is not a problem.\n\nBottom line: send to a serious referee. Require the author to align the statement of Theorem 3 with the body and to resolve the example contradictions. For a reading group, I would wait for a revision; for citation, I would wait as well.","headline":"New cπ1 machinery and a solid cutoff theorem, but the advertised Theorem 3 overstates what is proved and two examples flatly contradict themselves.","tokens_in":43377,"tokens_out":7136,"would_cite":false,"duration_ms":80489,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57N05","57M10","57M05"],"pacs":[],"model":"deepseek-v4-flash","headline":"For proper degree-one maps between non-compact surfaces, injectivity on loops at every end forces either a geometric kernel or a proper homotopy to a homeomorphism.","keywords":["geometric kernel","proper maps","non-compact surfaces","proper homotopy","proper fundamental group","ends of surfaces","degree-one maps","infinite-genus surfaces"],"falsifier":"A counterexample would be a proper degree-one map between oriented non-compact surfaces with injective induced map on ends and injective proper-π1 map along each end, but with no geometric kernel and not properly homotopic to a homeomorphism. The paper notes that the existence of a non-π1-injective, non-π1-surjective prime-degree self-map of the one-ended infinite-genus surface would produce such a counterexample by lifting, so constructing or ruling out such a map would settle the sharpness of the theorem.","tokens_in":42496,"feed_emoji":"","tokens_out":9817,"duration_ms":94249,"temperature":0.7,"pith_summary":"This paper asks when a proper map between non-compact surfaces must send some non-contractible simple loop to a null-homotopic loop, a geometric kernel. For closed surfaces every non-π1-injective map has such a kernel, but for open surfaces this can fail, and the paper shows that the failure is controlled by what happens at the ends. Using the proper fundamental group of an end, it proves that if a degree-one map is injective on end-spaces and on loops near every end along some ray, then outside a compact set the map is a finite-sheeted covering. Reducing to compact surfaces leaves exactly two possibilities: the map has a geometric kernel, or it is properly homotopic to a homeomorphism. The paper also gives end-based conditions under which infinitely many essential handles can be pinched, and it settles the planar case with a degree-zero dichotomy.","feed_headline":"End-wise injectivity forces a geometric kernel or a homeomorphism","feed_subtitle":"For degree-one proper maps of non-compact surfaces, injectivity on loops at every end leaves only two possibilities.","key_machinery":"The proper fundamental group of an end, based at a proper ray, records loops supported arbitrarily far out along that ray; its conjugacy-class version, cπ1, removes dependence on the chosen ray by encoding sequences of loops converging to the end up to free homotopy near the end. The load-bearing mechanism is the cut-off argument: end-wise π1-injectivity forces the map to behave as a finite-sheeted covering outside a compact set, after which the compact degree-one classification decides whether an essential handle is collapsed.","core_discovery":"The paper's central result is a cut-off theorem: for a proper map between non-compact surfaces, if for every end of the domain there is a representative ray along which the induced map on proper fundamental groups is injective, then the map can be properly homotoped so that, outside compact submanifolds, it is a finite-sheeted covering on each unbounded component. For a degree-one map whose induced map on ends is injective, this covering behavior has degree one, hence is a homeomorphism outside a compact piece. The compact piece then falls under the classical classification of degree-one maps between compact surfaces, whose two outcomes are a homeomorphism or a collapse of an essential compa","pith_inferences":["The end-wise π1-injectivity condition is likely not necessary for the conclusion: the paper's examples suggest that failure at a single end may be the only obstruction, and a natural test is whether a non-π1-injective degree-one self-map of a one-ended infinite-genus surface without a geometric kernel would violate the condition at that end.","The cπ1 singleton-preimage hypothesis behaves like a unique-lifting property at infinity; it could be rephrased in terms of profinite completions of end groups, raising the question of whether residual finiteness of surface groups makes such conditions automatically true for many quotient maps.","The planar dichotomy points to a broader pattern: degree-one proper maps between planar surfaces can fail to have geometric kernels, and the homology-level kernel found in the paper may be the strongest conclusion possible once end-allowability is dropped."],"forward_implications":["A proper degree-one map satisfying the end hypotheses is almost a covering: all failure of π1-injectivity is confined to a compact subsurface.","If such a map is not properly homotopic to a homeomorphism, it must collapse an essential compact bordered subsurface of genus at least one, so a geometric kernel exists.","For planar surfaces with at least three ends and injective induced map on ends, having a geometric kernel, having nontrivial kernel of π1, and having degree zero are equivalent.","Under the cπ1 singleton-preimage condition, a proper map can be homotoped to pinch essential handles, and infinitely many handles can be pinched exactly when the induced cπ1 map fails to be injective at some end.","Without end-allowability, degree-one planar maps can still produce a homology-level kernel, realized by a simple loop in the kernel of H1."],"supporting_citations":[{"why":"Introduces the proper fundamental group of an end, the invariant whose monomorphism condition drives the cut-off theorem.","marker":"[3]"},{"why":"The 3-manifold proper-homotopy classification whose proof strategy is adapted to surfaces.","marker":"[2]"},{"why":"The classification of π1-injective proper maps between boundaryless non-compact surfaces that this paper extends to bordered surfaces.","marker":"[7]"},{"why":"The classification of degree-one maps on compact surfaces: homotopic to a homeomorphism or to a pinch of an essential handle.","marker":"[8]"},{"why":"The compact-surface result that non-π1-injective maps admit geometric kernels, the base case after cutting off.","marker":"[14]"},{"why":"Shows a simple non-contractible curve is never a proper power of another loop, forcing local covering degree ±1 near ends.","marker":"[9]"},{"why":"Provides the degree-of-a-map tools used to turn finite-sheeted covering behavior at infinity into degree-one conclusions.","marker":"[10]"},{"why":"Gives the planar surface model of a sphere minus a totally disconnected set, used in the planar geometric-kernel theorems.","marker":"[21]"}],"fun_headline_variants":["End-injectivity on non-compact surfaces forces covering or kernel","Injective loops at every end imply geometric kernel or homeomorphism","Degree-one proper maps with injective ends: homeomorphism or kernel","Proper fundamental group injectivity yields geometric kernel","Non-compact maps: injective ends cut to coverings and cores"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The load-bearing premise is that, at every end of the domain, there is some ray escaping to that end along which every nontrivial sequence of loops far out remains nontrivial after mapping; if this fails at even one end, the covering behavior at infinity that drives the theorem need not hold.","fun_headline_variants_meta":{"raw":{"variants":["End-injectivity on non-compact surfaces forces covering or kernel","Injective loops at every end imply geometric kernel or homeomorphism","Degree-one proper maps with injective ends: homeomorphism or kernel","Proper fundamental group injectivity yields geometric kernel","Non-compact maps: injective ends cut to coverings and cores"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001297,"raw_usage":{"total_tokens":5071,"prompt_tokens":624,"completion_tokens":4447,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":368,"completion_tokens_details":{"reasoning_tokens":4371}},"tokens_in":368,"tokens_out":4447,"duration_ms":31445,"temperature":1.0,"reasoning_tokens":4371,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T14:35:33.124550+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A counterexample would be a proper degree-one map between oriented non-compact surfaces with injective induced map on ends and injective proper-π1 map along each end, but with no geometric kernel and not properly homotopic to a homeomorphism. The paper notes that the existence of a non-π1-injective, non-π1-surjective prime-degree self-map of the one-ended infinite-genus surface would produce such a counterexample by lifting, so constructing or ruling out such a map would settle the sharpness of the theorem.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the proper fundamental group of an end, the invariant whose monomorphism condition drives the cut-off theorem."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The 3-manifold proper-homotopy classification whose proof strategy is adapted to surfaces."},{"cited_title":"$\\pi_1$-injective proper maps between non-compact surfaces","cited_arxiv_id":"2405.09824","evidence_quote":"The classification of π1-injective proper maps between boundaryless non-compact surfaces that this paper extends to bordered surfaces."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The classification of degree-one maps on compact surfaces: homotopic to a homeomorphism or to a pinch of an essential handle."},{"cited_title":"The simple loop conjecture","cited_arxiv_id":null,"evidence_quote":"The compact-surface result that non-π1-injective maps admit geometric kernels, the base case after cutting off."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows a simple non-contractible curve is never a proper power of another loop, forcing local covering degree ±1 near ends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the degree-of-a-map tools used to turn finite-sheeted covering behavior at infinity into degree-one conclusions."},{"cited_title":"On the classification of noncompact surfaces","cited_arxiv_id":null,"evidence_quote":"Gives the planar surface model of a sphere minus a totally disconnected set, used in the planar geometric-kernel theorems."}],"review_version":1}