{"id":"1b68009d-d036-4366-8a91-93591f30139c","arxiv_id":"2411.13233","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A fiberwise Nielsen periodic number NBP_n(f) is defined, shown to bound the minimal number of n-periodic points up to homotopy over B, and computed explicitly for S^1-bundles over S^1.","lead":"This paper defines a version of the Nielsen periodic number for maps that respect the layers of a fibration, counting periodic points in a way that homotopies over the base space cannot destroy. The invariant could be used to determine which periods are unavoidable in fiber-preserving dynamical systems, such as maps of tori over the circle.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 5.3's proof uses the wrong divisor sum: the depth-m count must subtract proper divisors of m, not of n, so the equality A_n(f)=NBP_n(f) is not established as written.","rationale":"I read the paper as establishing a fiberwise Nielsen periodic number NBP_n(f) and deriving its main lower-bound and computability properties. The most load-bearing concern is not the path-connectedness of the fiber, which the manuscript states at the start of Section 2 if the 'all spaces' clause includes the fiber Y in Y→E→B; under that reading every fiber is path-connected because B is path-connected. The true soft spot is the divisor-summation error in the proof of Theorem 5.3. Since A_m(f) is defined by subtracting sums over proper divisors of m, the proof's subtraction over divisors of n is generally unequal to A_m(f), and the printed equality is literally false in the paper's own torus example. This is a correctable local flaw rather than a disproof of the statement, but it is the step on which the central equality A_n(f)=NBP_n(f) depends, and it propagates to Corollary 5.5 and Proposition 5.8. The reader's verdict was already CONDITIONAL, and this stress-test confirms that the manuscript needs a corrected summation or a detailed justification before the central computation can be accepted as written.","tokens_in":16685,"tokens_out":14210,"duration_ms":146717,"concrete_test":"Recompute the cardinality of the subset of RB(f^n; x0, n(ω)) of depth exactly m, for each proper divisor m of n, using injectivity of γ_{m,n}: the number of images of irreducible classes of RB(f^m; x0, m(ω)) is A_m(f) = NB(f^m) − Σ_{k|m, k<m} A_k(f), and this expression must replace the printed NB(f^m) − Σ_{j|n, j<n} A_j(f) in Theorem 5.3. A concrete numerical test is the map f_{1,s} of Proposition 5.8 with n=6, m=2 and s ≠ 0: the printed formula gives NB(f^2) − A_1 − A_2 − A_3 = 2|s| − |s| − |s| − 2|s| = −2|s|, an impossibility, while the corrected index set gives NB(f^2) − A_1 = |s| = A_2. Verify the corrected identity for n=6, m=2,3 and for n=12, m=2,3,4,6.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central equality A_n(f)=NBP_n(f) in Theorem 5.3 rests on a counting step whose displayed summation is over the wrong index set. The proof says: 'the cardinality of Reidemeister classes of f^n with depth m < n is precisely NB(f^m) − Σ_{j|n, j<n} A_j(f) = A_m(f).' But Definition 5.2 defines A_m(f) = NB(f^m) − Σ_{k|m, k<m} A_k(f). The index set in the proof must be the proper divisors of m, not the proper divisors of n. The displayed equality is false in general: for n=6 and m=2, the right-hand side NB(f^2) − A_1(f) − A_2(f) − A_3(f) is not A_2(f); in the S^1-bundle example of Proposition 5.8 it equals −2|s| while A_2(f)=|s|. This is not merely a notational slip, because the conclusion that the number of irreducible essential classes of f^n equals NB(f^n) − Σ_{k|n, k<n} A_k(f) depends on counting, for each proper divisor m, exactly the classes of depth m. That conclusion is likely recoverable by replacing the sum over divisors of n with the sum over proper divisors of m, but the proof as printed does not supply that corrected argument.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a Nielsen-type periodic number NBP_n(f) for fiber-preserving self-maps f:E→E over a base B, extending the classical Nielsen periodic number NP_n. It defines algebraic and geometric Reidemeister classes over B, studies orbits and depths of periodic classes, and proposes a lower bound NBP_n(f) ≤ min{#π0(P_n(g)) | g ∼_B f}. Under an n-toral hypothesis and the assumption NB(f^m)=RB(f^m)≠0 for all m|n, it claims A_n(f)=NBP_n(f), where A_n is defined by Möbius inversion over the divisors of n. The paper then applies this to S^1-bundles over S^1 and derives an Euler-phi-like formula in Proposition 5.8.","tokens_in":16955,"tokens_out":41496,"duration_ms":420395,"significance":"If the central definition and theorems were correct, the paper would contribute a natural Nielsen periodic number over a base space, with a homotopy-invariant lower bound and explicit computations for S^1-bundles. The algebraic framework of Reidemeister classes over B and the use of standard fibration and index theory from [3] and [14] are sensible, and the paper has no fitted parameters or circular dependencies. However, the main invariant as defined does not satisfy the advertised lower bound, and the proof of the key equality in Theorem 5.3 has a false divisor summation. These are load-bearing issues that currently prevent acceptance.","major_comments":[{"comment":"The definition NBP_n(f)=n×O_n(f), where O_n(f) counts orbits of irreducible essential Reidemeister classes, is not compatible with the claimed lower bound. Consider the paper's own example setup with T=S^1×S^1, p(x,y)=y, and f(x,y)=(xy,y); this is f_{1,1} in Proposition 5.8. For n=2, the paper's computation gives R_{S^1}(f^2)=2 and R_{S^1}(f)=1, so there is exactly one irreducible essential class of f^2. Because the fiber map is trivial on π1 of the fiber at the fixed fiber y=1, the induced action [f^ω] on R_{S^1}(f^2) is the identity, so this irreducible class forms a singleton orbit. Hence O_2(f)=1 and Definition 4.1 gives NBP_2(f)=2. But P_2(f)=S^1×{-1}, which is connected, so #π0(P_2(f))=1. This contradicts Proposition 4.2(ii). The proof of Proposition 3.3, which asserts that x and f^l(x) lie in different path components of Fix(f^d), is also false in this example: x and f(x) lie in the same connected fiber S^1×{-1}. The invariant needs to be redefined—likely by counting irreducible essential classes rather than multiplying the number of class-orbits by n.","section":"Definition 4.1 and Proposition 4.2(ii)"},{"comment":"The displayed equality in the proof of Theorem 5.3 is false as written: the proof states that the number of Reidemeister classes of f^n with depth m<n is NB(f^m)−Σ_{j|n, j<n}A_j(f)=A_m(f). Definition 5.2 defines A_m(f)=NB(f^m)−Σ_{k|m, k<m}A_k(f), so the summation must be over the proper divisors of m, not of n. For n=6 and m=2, the displayed right-hand side NB(f^2)−A_1(f)−A_2(f)−A_3(f) is not equal to A_2(f). The subsequent conclusion NBP_n(f)=NB(f^n)−Σ_{k|n,k<n}A_k(f)=A_n(f) depends on this incorrect counting step. A corrected argument can likely be supplied by considering the fixed-point set of [f^ω]^d on R_B(f^n) and applying Möbius inversion, but that argument is not present in the manuscript.","section":"Theorem 5.3, proof"},{"comment":"The proof of Proposition 5.7 establishes only the injectivity of [γ_{k,n}], not condition (i) of Definition 5.1, which requires d(<[α]_m>)=l(<[α]_m>) for every m|n. This is not a minor omission: for the map f_{1,1} of the previous comment, condition (i) fails for m=2, since the unique irreducible class of f^2 has depth 2 but orbit length 1. Thus Proposition 5.7's conclusion that such f is n-toral is false without an additional hypothesis, and Proposition 5.8 cannot invoke Theorem 5.3 for f_{1,s} merely from the equality NB(f^m)=RB(f^m).","section":"Proposition 5.7"}],"minor_comments":[{"comment":"The standing assumption that 'all spaces are path-connected' should explicitly state that every fiber of p is path-connected, not only E and B. The path ω in Definition 2.1 exists only if x0 and f(x0) lie in the same path component of the fiber.","section":"Section 2, opening"},{"comment":"Reference [14] is listed as 'Springer-Verlag, 1918'; the intended year is presumably 1978.","section":"References"},{"comment":"The final sentence of Proposition 3.3 claims that the points x, f^l(x), ... belong to different path components of Fix(f^d). This is not established by the cited Proposition 2.8 and is false for the f_{1,1} example discussed above; the proof needs to be reworked or the statement corrected.","section":"Proposition 3.3"}],"recommendation":"major_revision","confidential_remarks":"The paper has a promising framework, but the central definition and its lower bound need reworking. I would not publish the current version. The f_{1,1} counterexample is within the paper's own scope and should be addressed directly. If the authors redefine NBP_n appropriately and repair the proof of Theorem 5.3, the corrected results may be publishable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a competent, incremental extension of Heath–Piccinini–You periodic Nielsen theory to maps over a base, using the Gonçalves–Koschorke fiberwise Nielsen theory. The new invariant NBP_n(f) is natural, homotopy invariant over B, and does what a Nielsen periodic number should. But Theorem 5.3, the main computational result, has a proof gap: the counting sum runs over the wrong set of divisors. As printed, the equality is false. That needs fixing.\n\nCredit where due: the paper defines algebraic periodic Reidemeister classes over B, gives them a depth and length, and packages them into NBP_n(f). The classical number is recovered when B is a point, which is the right sanity check. The homotopy invariance argument in Proposition 4.2(i) is mostly careful — the path homotopy over B is worked out in some detail. The torus computation in Proposition 5.8 gives a clean closed formula, the kind of concrete result that makes the theory usable. The n-toral and NB=RB hypotheses are stated plainly as assumptions, not smuggled in.\n\nSoft spots, in proportion. The Theorem 5.3 proof: Definition 5.2 sets A_m(f) = NB(f^m) − Σ_{k|m,k<m} A_k(f). But the proof claims the number of Reidemeister classes of f^n with depth m < n is NB(f^m) − Σ_{j|n, j<n} A_j(f) = A_m(f). The index set is wrong; it should be proper divisors of m, not of n. For n=6, m=2 the displayed expression is not A_2(f). The conclusion is probably recoverable — depth-m classes in f^n correspond to irreducible classes of f^m under the injective γ_{m,n}, and those are counted by A_m(f) — but the printed proof does not say that. Since this underlies the inclusion-exclusion formula and the torus example, the paper needs a corrected argument, not just a typo fix.\n\nSecond, the standing assumption that all spaces are path-connected is ambiguous about the fiber. Definition 2.1 needs a path ω inside the fiber; if fibers are not path-connected, the construction collapses. Path-connected E and B do not force path-connected fibers. The authors should say explicitly that the fiber is path-connected (or adjust the definition).\n\nMinor: the lower-bound proof in Proposition 4.2(ii) is a bit compressed but the idea is standard and likely fine.\n\nBottom line: the paper is a solid incremental contribution, not a breakthrough. The right audience is people in Nielsen theory or fiberwise fixed point theory. It deserves a serious referee; the main theorem is plausible but unproved as written. Send it out, ask for the divisor-sum correction and a clarification of the fiber assumption.","headline":"A natural fiberwise Nielsen periodic number with a real, fixable counting error in the main theorem; worth refereeing.","tokens_in":17523,"tokens_out":5817,"would_cite":false,"duration_ms":56081,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["55M20","55R10","37C25"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper introduces a Nielsen-type periodic number $NBP_n(f)$ for fiber-preserving maps over a base space $B$, proves it is homotopy invariant over $B$, and computes it for $S^1$-bundles over $S^1$.","keywords":["Nielsen periodic number","Reidemeister classes over B","fiber-preserving maps","periodic points","fibrations","homotopy invariance","minimal periods","torus bundles"],"falsifier":"Take $B=S^1$ and let $E$ be the mapping torus of the transposition of a two-point fiber, so $E$ and $B$ are path-connected but the fiber is $S^0$. A fiber map over $B$ that swaps the two points of a fiber has no path $\\omega$ in that fiber from $x_0$ to $f(x_0)$, so Definition 2.1, Theorem 2.6 and Proposition 4.2 cannot be formulated; this example marks exactly where the connected-fiber premise is load-bearing.","tokens_in":16426,"feed_emoji":"🔁","tokens_out":14800,"duration_ms":133535,"temperature":0.7,"pith_summary":"This paper extends the classical Nielsen periodic number from ordinary self-maps to fiber-preserving maps over a fixed base space $B$. Its central object is $NBP_n(f)$, defined as $n$ times the number of irreducible essential periodic orbits of Reidemeister classes over $B$, and the paper proves that this number is invariant under homotopies over $B$ and is a lower bound for the minimal number of periodic-point orbits among all maps homotopic over $B$ to $f$. When $B$ is a single point, $NBP_n(f)$ reduces to the classical Nielsen periodic number $NP_n(f)$. The paper also isolates a class of $n$-toral maps over $B$ for which $NBP_n(f)$ equals the inclusion-exclusion count $A_n(f)$, and computes $NBP_n(f)$ explicitly for $S^1$-bundles over $S^1$.","feed_headline":"New Nielsen number counts periodic orbits over a base space","feed_subtitle":"It matches classical Nielsen periodic numbers when B is a point and yields torus-bundle formulas.","key_machinery":"The load-bearing mechanism is the fibrewise Reidemeister equivalence. For a point $x_0$ and a path $\\omega$ in the fiber $p^{-1}(p(x_0))$ from $x_0$ to $f(x_0)$, the action $[c]*_B[\\theta]=[\\tilde H(1,s)*\\omega^{-1}]$ is defined by lifting the base path $p(c)$ to a homotopy whose vertical sides are $\\theta*\\omega$ and $f(c)$; equivalence classes are the algebraic Reidemeister classes over $B$, and by Theorem 2.6 they are in bijection with the path-components of the space $E_B(f)$ of pairs $(x,\\alpha)$ with $\\alpha$ a path in the fiber from $x$ to $f(x)$. The periodic theory is carried by the induced map $[f_\\omega]$ acting on classes of $f^n$, the maps $[\\gamma_{m,n}]$ for $m|n$, and the notions of length, depth and irreducibility; $O_n(f)$ counts the irreducible essential orbits, and $NBP_n(f)$ multiplies this count by $n$. This setup is what makes homotopy invariance over $B$ and the inclusion-exclusion formula possible.","core_discovery":"The central claim is that the Nielsen periodic count can be carried out fibrewise and behaves like its classical counterpart. For a fibration $Y\\to E\\to B$ and a fiber map $f:E\\to E$ over $B$, the paper defines algebraic Reidemeister classes over $B$ for $f^n$ via a twisted action of $\\pi_1(E,x_0)$ on $\\pi_1(Y,x_0)$, using a path $\\omega$ in the fiber from $x_0$ to $f(x_0)$. It shows these classes are in bijection with the path-components of the fiber space $E_B(f^n)$, so the geometric fixed-point classes over $B$ inject into them. The resulting number $NBP_n(f)=n\\,O_n(f)$, where $O_n(f)$ counts irreducible essential periodic orbits of these classes, is proved to be a homotopy invariant over $B$ and to satisfy $NBP_n(f)\\le\\min\\{\\#\\pi_0(P_n(g)):g\\sim_B f\\}$. Under the $n$-toral condition and the assumption that $0\\neq NB(f^m)=RB(f^m)$ for every $m|n$, the paper proves $A_n(f)=NBP_n(f)$, and for $T=S^1\\times S^1$ over $S^1$ with $f$ homotopic over $S^1$ to $f_{1,s}$, $s\\neq0$, it yields $N_{S^1}P_n(f)=|s|\\,p_1^{\\alpha_1-1}\\cdots p_l^{\\alpha_l-1}(p_1-1)\\cdots(p_l-1)$.","pith_inferences":["The same depth-and-length machinery may yield closed formulas for other fiber bundles, such as nilmanifold or infranilmanifold bundles, where the fundamental-group action on the fiber is more complicated than in the $S^1$ case.","If a sharpening theorem over $B$ were established, the inequality in Proposition 4.2 would become an equality; the paper does not prove such a result.","The conditions in Theorem 5.3 might be relaxed to a weaker essentiality assumption, which would make $NBP_n(f)$ computable for a larger class of fiber maps."],"forward_implications":["If $NBP_n(f)\\neq 0$, then $n$ belongs to the homotopy-invariant period set $HBPer(f)$, so the invariant is a tool for detecting periods that persist under homotopies over $B$.","When $B$ is a point, $NBP_n(f)$ collapses to the classical Nielsen periodic number $NP_n(f)$, recovering the ordinary theory as a special case.","For $S^1$-bundles over $S^1$, the closed formula $N_{S^1}P_n(f)=|s|\\,p_1^{\\alpha_1-1}\\cdots(p_l-1)$ turns the invariant into an explicit count in terms of one integer $s$.","The inequality $NBP_n(f)\\le\\min\\{\\#\\pi_0(P_n(g)):g\\sim_B f\\}$ converts the algebraic count into a lower bound on periodic-point orbits for every map homotopic over $B$ to $f$."],"supporting_citations":[{"why":"supplies the Nielsen theory over B, the index of fixed-point classes over B, and Theorem 1.3 giving N_B = R_B = MCF for S1-bundles used in the torus computation.","marker":"[3]"},{"why":"defines the classical Nielsen periodic number NP_n whose formula NBP_n matches when B is a point and whose orbit-depth framework is adapted here.","marker":"[5]"},{"why":"provides the homotopy lifting results (I.7.16, I.7.18) used to define the Reidemeister action over B and to prove the canonical bijection with path-components of E_B(f).","marker":"[14]"},{"why":"gives the normal form f_{r,s} for maps over S1 of S1-bundles, which underlies Propositions 5.7 and 5.8.","marker":"[4]"},{"why":"supplies the fiber homotopy equivalence E_B(f^n) to E_B(g^n) preserving essentiality, used in the homotopy-invariance proof of NBP_n.","marker":"[12]"}],"fun_headline_variants":["Nielsen periodic count generalized to fiber bundles","New periodic number tracks orbits over base spaces","Fiberwise Nielsen number counts periodic classes","Over a base space, a new Nielsen periodic invariant","Periodic points counted over a fibration's base"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Every fiber $Y=p^{-1}(b)$ of the fibration is assumed path-connected, because Definition 2.1 needs a path $\\omega$ lying entirely inside a single fiber from $x_0$ to $f(x_0)$; if a fiber is disconnected, such a path need not exist and the Reidemeister classes over $B$ are undefined.","fun_headline_variants_meta":{"raw":{"variants":["Nielsen periodic count generalized to fiber bundles","New periodic number tracks orbits over base spaces","Fiberwise Nielsen number counts periodic classes","Over a base space, a new Nielsen periodic invariant","Periodic points counted over a fibration's base"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000627,"raw_usage":{"total_tokens":2909,"prompt_tokens":960,"completion_tokens":1949,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":576,"completion_tokens_details":{"reasoning_tokens":1879}},"tokens_in":576,"tokens_out":1949,"duration_ms":12589,"temperature":1.0,"reasoning_tokens":1879,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:42:19.143659+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $B=S^1$ and let $E$ be the mapping torus of the transposition of a two-point fiber, so $E$ and $B$ are path-connected but the fiber is $S^0$. A fiber map over $B$ that swaps the two points of a fiber has no path $\\omega$ in that fiber from $x_0$ to $f(x_0)$, so Definition 2.1, Theorem 2.6 and Proposition 4.2 cannot be formulated; this example marks exactly where the connected-fiber premise is load-bearing.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"supplies the Nielsen theory over B, the index of fixed-point classes over B, and Theorem 1.3 giving N_B = R_B = MCF for S1-bundles used in the torus computation."},{"cited_title":"Heath, R","cited_arxiv_id":null,"evidence_quote":"defines the classical Nielsen periodic number NP_n whose formula NBP_n matches when B is a point and whose orbit-depth framework is adapted here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the homotopy lifting results (I.7.16, I.7.18) used to define the Reidemeister action over B and to prove the canonical bijection with path-components of E_B(f)."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"gives the normal form f_{r,s} for maps over S1 of S1-bundles, which underlies Propositions 5.7 and 5.8."},{"cited_title":"Koschorke; Nielsen coincidence theory in arbitrary codimensions , 598, J","cited_arxiv_id":null,"evidence_quote":"supplies the fiber homotopy equivalence E_B(f^n) to E_B(g^n) preserving essentiality, used in the homotopy-invariance proof of NBP_n."}],"review_version":1}