{"id":"45d036b5-3b26-451a-9d0e-4e1f581bd201","arxiv_id":"1908.05269","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Exotic Mazur manifolds exist, and the knot Floer invariant ν is shown to be an invariant of knot traces, yielding counterexamples to a conjecture in Kirby's problem list.","lead":"This paper constructs the first pairs of Mazur manifolds that are homeomorphic but not diffeomorphic, settling an open question in 4-manifold topology. The key tool is a proof that the knot Floer invariant ν is almost always an invariant of knot traces, the 4-manifolds built by attaching a 2-handle to a ball along a knot.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No internal inconsistency found; the conditional verdict stands because the central claim depends on SnapPy/Sage hyperbolicity and volume checks plus a monotone Dehn-filling step that the paper outsources to [18] and does not certify in text.","rationale":"The paper's central argument is coherent: the handle calculus gives homeomorphic Mazur manifolds, the trace invariance of ν is proved internally with the stated n < 0 exception, and the twist inequality plus Levine's formula supplies the smooth obstruction. I do not see an internal inconsistency in the main line from Theorem 1.4 and Theorem 2.11 to Theorem 1.1. The genuine soft spot is the step from trace exotica to Mazur exotica: it requires the JSJ hypothesis on the exterior E, and for the infinite family this is established by a finite set of SnapPy/Sage volume and hyperbolicity computations plus an asymptotic monotonicity argument. The paper reports the numerical values but does not reproduce the certificates or give the explicit threshold N, and the full log is in an external data repository. That is a load-bearing dependency because a failure there would invalidate Theorem 2.7's hypotheses for infinitely many of the constructed pairs. The reader identified the same point as the weakest assumption, and the conditional verdict already reflects it. The reliance of Theorem 1.5 on unpublished work of Ozsvath-Szabo is a separate concern but does not affect the central claim, so I would not change the reader's verdict.","tokens_in":24968,"tokens_out":45912,"duration_ms":468741,"concrete_test":"Independently rerun SnapPy's certified hyperbolicity verification (with interval arithmetic) and volume computations for S^3_0(C)\\K_1, S^3_0(C)\\(K ∪ α), and V0(P), starting from the triangulations in [18]; then recompute the volumes of S^3_0(C)\\K_m for m = 2 through a moderately large M by performing -1/(m-1) surgery on α, and confirm that the filled manifolds are certified hyperbolic, that their volumes increase monotonically, and that all volumes remain bounded away from vol(V0(P)) by a clear margin. If the certification fails or a volume dip occurs, the infinite-family hypothesis preceding Theorem 2.7 is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing bridge for Theorem 1.1 is Theorem 2.7: a diffeomorphism of the Mazur pair yields diffeomorphic n-traces, provided the exterior E of K in S^3_0(C) is not the solid torus and has no JSJ component homeomorphic to V0(P). The verification that the infinite family L_m = K_m ∪ C satisfies this hypothesis is not carried out in the paper: for m = 1 and for the two-cusped link K ∪ C ∪ α the hyperbolicity and volume computations are reported only as SnapPy/Sage estimates (vol ≈ 9 and vol ≈ 11, with vol(V0(P)) < 4), and the data are relegated to the external repository [18]. For m > N the argument invokes Thurston's hyperbolic Dehn surgery theorem and [38, Theorem 1A] to assert that volume increases monotonically to ≈ 11, but no explicit N is given and the monotonicity claim is not verified against the specific cusp basis used by SnapPy. If any of these computations is not certified, or if the monotonicity step is applied outside its hypotheses, the JSJ uniqueness of V0(P) in the boundary Y is not established, and the bridge from exotic knot traces to exotic Mazur manifolds collapses. The dependence of Theorem 1.5 on the unpublished [40] is real but secondary, since it does not affect the proof of Theorem 1.1.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs pairs of Mazur manifolds, proves they are homeomorphic but not diffeomorphic (Theorem 1.1), and introduces a new smooth invariant of knot traces derived from the Heegaard Floer concordance invariant ν (Theorem 1.4). It also shows, conditional on forthcoming work of Ozsváth and Szabó [40], that τ and ε are not zero-trace invariants (Theorem 1.5), and derives shake genus bounds from ν (Theorem 1.6). The main technical engine is a careful mapping-cone argument in §4 that establishes trace invariance of ν, combined with a handle-theoretic bridge (Theorem 2.7) that converts Mazur diffeomorphisms into trace diffeomorphisms for links with certain JSJ hypotheses.","tokens_in":25235,"tokens_out":31000,"duration_ms":283844,"significance":"If Theorem 1.1 holds, it provides the first examples of exotic smooth structures on Mazur manifolds, a central question in 4-manifold topology, and it yields the first counterexamples to the uniqueness conjecture for S^1×S^2 surgeries on integer homology spheres (Corollary 1.3). The trace invariance of ν is a genuinely new and computable tool for studying knot traces, and the paper offers reproducible computational data in the repository [18]. The mapping-cone proof of Theorem 1.4 is a serious, self-contained derivation from established Heegaard Floer theory, and the paper is careful to flag the dependence of Theorem 1.5 on the unpublished [40]. These are significant strengths. However, the proof of the infinite family in Theorem 1.1 relies on a volume-monotonicity claim that is not fully certified, which prevents the main theorem from being accepted at face value.","major_comments":[{"comment":"The verification that the infinite family L_m satisfies the hypotheses preceding Theorem 2.7 depends on the assertion that -1/(m-1)-surgery on α produces a hyperbolic manifold whose volume increases monotonically with m for m>N, citing Thurston's hyperbolic Dehn surgery theorem and [38, Theorem 1A]. No explicit N is given, and the monotonicity is not checked against the specific cusp basis used in the SnapPy/Sage computations. If the monotonicity fails for some large m, the conclusion that S^3_0(C)\\K_m is not diffeomorphic to V0(P) is not established, and the JSJ uniqueness step in Theorem 2.7 would not apply. The authors should supply a rigorous proof of the monotonicity claim or provide an explicit N together with certified computations for all m>N.","section":"§2.3, Proof of Theorem 1.1"},{"comment":"Theorem 1.5, stating that τ and ε are not zero-trace invariants, depends crucially on the unpublished equivalence in [40] between the bordered invariants τ, ε, ν and their standard counterparts. Although the abstract and the proof mention this dependence, the theorem is stated unconditionally in §4.3. The authors should either state the theorem as conditional on [40], or, if the result is considered established, provide a verifiable reference. This is a load-bearing issue for Theorem 1.5, though not for Theorem 1.1.","section":"§4.3, Theorem 1.5"},{"comment":"The formula ν(P(K)) = ν(K)+1 if ν(K)=τ(K)>0 and ν(P(K))=ν(K) otherwise is stated with the proof left to the reader. This result is used directly in the proof of Theorem 2.11, which is a key step in establishing Theorem 1.1. Since the central theorem depends on this calculation, a complete proof or a detailed derivation from Theorem 2.8 and the properties of ν should be included.","section":"§2.3, Corollary 2.9"}],"minor_comments":[{"comment":"Reference [13] is misattributed: the citation 'A Donald, Embedding Seifert manifolds in S4' is not the source of Donaldson's theorem on intersection forms used in §4.2. The correct reference is S.K. Donaldson, 'An application of gauge theory to four-dimensional topology', J. Differential Geom. 18 (1983) 279-315.","section":"References"},{"comment":"In the proof of Proposition 4.2, the phrase 'unless s = ν(K) = 0' should read 'unless s = ε(K) = 0' to match Lemma 4.1; the intended meaning is clear because ε(K)=0 implies ν(K)=0, but the notation is inconsistent as written.","section":"§4.1, Proposition 4.2 proof"},{"comment":"The paper reports hyperbolicity and volume computations from SnapPy and Sage only as approximate values (vol ≈ 9, vol ≈ 11) and refers to the external repository [18] for documentation. Including the exact link diagrams and the specific commands or scripts in the text would improve reproducibility and allow the reader to verify the JSJ hypotheses without accessing the repository.","section":"§2.1, SnapPy/Sage computations"},{"comment":"The handle calculus in Figure 3 is used repeatedly in the proof of Proposition 2.2, but the individual moves are not annotated. Adding step-by-step explanations or labels for each move would make the argument significantly easier to follow.","section":"§2.2, Figure 3"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely to be a significant contribution if the computational gap in the infinite family is closed. I would recommend that the handling editor seek verification of the volume-monotonicity step, ideally by an expert in hyperbolic Dehn surgery or by requiring the authors to provide an explicit N and certified computations. The dependence on [40] for Theorem 1.5 is clearly flagged, but the theorem should be phrased accordingly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Start here: this is a genuine advance, not a repackaging. The paper produces the first exotic Mazur manifolds (homeomorphic, not diffeomorphic), and it does so with a new invariant: ν from knot Floer homology is shown to be invariant under diffeomorphisms of oriented knot traces, with one explicitly excluded edge case. That trace invariance (Theorem 1.4) is the technical heart, and the mapping cone argument in §4 looks correct to me. The handle calculus establishing homeomorphism of the Mazur pairs is also convincing, and the resolution of the S^1×S^2 surgery question from Kirby's list follows cleanly.\n\nThe paper is honest about its own load-bearing assumptions. Theorem 1.5 depends on the unpublished Ozsváth–Szabó equivalence [40]; the authors say so. The infinite family L_m satisfies the JSJ hypothesis only via SnapPy/Sage hyperbolicity and volume estimates, with data in an external repository. No explicit N is given for the monotone volume increase step. These are the genuine soft spots, and they are exactly where I would push a referee to demand a fuller record: either include the [40] identification or isolate the proof, and report the SnapPy/Sage computations in enough detail that the volume and hyperbolicity claims are reproducible rather than taken on faith from [18].\n\nSome smaller things: Corollary 2.9 leaves a proof to the reader that is genuinely routine from the surrounding relations, and the exception in Theorem 1.4 for n<0 and ν-values {0,1} is stated rather than resolved. Neither threatens the examples in the paper. I also note that the author-overlap citations (Mark–Tosun, Miller–Piccirillo, Piccirillo) are for auxiliary lemmas; the central claim does not rest on them. The computational dependence is a presentation gap, not a circular argument.\n\nWho should read this: anyone working in 4-manifold topology, especially handlebody and knot trace people. A serious referee will be needed; the paper deserves peer review and should be conditionally accepted with the expectation that the computational evidence and the [40] dependence be tightened. I would bring it to our reading group.","headline":"First exotic Mazur manifolds, with a new ν-based trace invariant; the central argument is sound, and the main weaknesses are heavy reliance on computational checks and an unpublished theorem, both explicitly flagged by the authors.","tokens_in":25811,"tokens_out":2343,"would_cite":true,"duration_ms":23994,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57K40","57K18","57K10"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that infinitely many Mazur manifolds—the simplest contractible 4-manifolds after the 4-ball—come in homeomorphic but non-diffeomorphic pairs, and that the knot Floer invariant ν detects them.","keywords":["exotic 4-manifolds","Mazur manifolds","knot traces","knot Floer homology","concordance invariants","Dehn surgery","shake genus","hyperbolic 3-manifolds"],"falsifier":"Run the paper's computer pipeline on the exterior of $K_m$ in $S^3_0(C)$ for each $m$ and inspect the JSJ torus decomposition: if some $m$ yields the solid torus or a component homeomorphic to $V_0(P)$, the bridge from Mazur diffeomorphisms to trace diffeomorphisms breaks for that $m$. Alternatively, find knots $K$ and $K'$ with diffeomorphic $n$-traces for some $n<0$ and $\\{\\nu(K),\\nu(K')\\}=\\{0,1\\}$; that would realize the one exceptional case allowed by Theorem 1.4 and show $\\nu$ is not a trace invariant for all framings.","tokens_in":24712,"feed_emoji":"🌀","tokens_out":14004,"duration_ms":125662,"temperature":0.7,"pith_summary":"This paper proves that Mazur manifolds—compact, contractible 4-manifolds built from one 1-handle and one 2-handle—admit exotic smooth structures: there are infinitely many pairs that are homeomorphic but not diffeomorphic. The proof converts this 4-dimensional question into a question about knot traces, the manifolds obtained by attaching an n-framed 2-handle to the 4-ball along a knot, and shows that the knot Floer concordance invariant ν is an invariant of the smooth oriented trace for essentially all framings. This yields a computable obstruction to trace diffeomorphisms that can work even when genus-function methods give no information. As corollaries, the paper produces integer homology spheres containing two distinct knots with $S^1\\times S^2$ surgeries, and shows (modulo a forthcoming identification of bordered invariants) that the concordance invariants τ and ε are not zero-trace invariants.","feed_headline":"Exotic Mazur manifolds exist in infinite families","feed_subtitle":"A knot Floer invariant distinguishes homeomorphic 4-manifolds when their smooth structures differ.","key_machinery":"The load-bearing object is the knot trace $X_n(K)=B^4$ with an $n$-framed 2-handle attached along $K$, together with the knot Floer invariant $\\nu(K)$. The technical engine is the mapping cone formula for Heegaard Floer homology of integer surgeries: for each spin$^c$ structure $\\mathfrak{s}$ on the surgery cobordism, the induced map $F_{\\mathfrak{s}}$ is shown to be zero or nonzero according to whether $|\\langle c_1(\\mathfrak{s}),\\sigma\\rangle|$ is above or below $2\\nu(K)-n$, with the borderline case governed by $\\epsilon(K)$. This converts the diffeomorphism type of a trace into the pattern of which cobordism maps are nonzero, and hence into the value of $\\nu$. The second piece is the satellite pair $P_n$ and $Q_n$: the patterns are concordant in the solid torus, so $Q_n(K)$ stays concordant to $K$ while $P_n(K)$ raises $\\nu$, and the two handle attachments are homeomorphic through a cork twist. The bridge from Mazur manifolds to traces is Theorem 2.7, which uses the JSJ decomposition and a hyperbolicity/volume check to show that any diffeomorphism of the Mazur pair induces a diffeomorphism of the associated traces.","core_discovery":"The central claim is that homeomorphism does not determine diffeomorphism type among Mazur manifolds, and that the difference is visible to knot Floer homology. For each link in an infinite family $L_m=K_m\\cup C$, the paper constructs two Mazur manifolds $W_{L_m,n}$ and $W'_{L_m,n}$ by attaching an n-framed 2-handle along the satellite knots $P_n(K_m)$ and $Q_n(K_m)$; the two are homeomorphic because the two attachment patterns differ by a cork twist, which preserves homeomorphism type. The diffeomorphism obstruction is Theorem 1.4: if the oriented knot traces $X_n(K)$ and $X_n(K')$ are diffeomorphic, then $\\nu(K)=\\nu(K')$, except possibly when $n<0$ and $\\{\\nu(K),\\nu(K')\\}=\\{0,1\\}$. Since $\\nu(P_n(K_m))=\\nu(K_m)+1$ while $\\nu(Q_n(K_m))=\\nu(K_m)$, the traces are not diffeomorphic whenever $\\nu(K_m)=\\tau(K_m)>0$, and Theorem 2.7 converts this into non-diffeomorphism of the Mazur manifolds. The passage from Mazur diffeomorphisms to trace diffeomorphisms uses 3-manifold topology—JSJ decompositions and hyperbolicity—to guarantee that a boundary diffeomorphism preserves the decomposition into the exterior of $K$ and the surgery solid torus.","pith_inferences":["Because $\\nu$ is computable in practice, the same trace-invariance criterion should be usable to search for exotic pairs among other satellite patterns and framings, not just the $P_n/Q_n$ family.","If the exceptional case $n<0$ with $\\{\\nu(K),\\nu(K')\\}=\\{0,1\\}$ is ever realized, it would produce the first counterexamples to full trace invariance of $\\nu$; no such examples are currently known.","The strategy of detecting exotic 4-manifolds by first passing to knot traces and then to a concordance invariant suggests that other mapping-cone-defined concordance invariants may also be trace invariants, yielding further computable obstructions.","The only non-theorem step in producing the infinite family is the computer-assisted hyperbolicity and volume verification; a closed-form proof for all $m$ would remove the computational dependence."],"forward_implications":["Infinitely many Mazur manifolds come in exotic pairs; earlier examples of exotic contractible 4-manifolds required more complicated handle structures.","The invariant $\\nu$ is a smooth invariant of oriented knot traces for $n\\ge 0$, and for $n<0$ except possibly when $\\{\\nu(K),\\nu(K')\\}=\\{0,1\\}$, giving a computable way to distinguish homeomorphic traces.","There are infinitely many irreducible integer homology spheres, each containing two distinct knots whose zero-surgeries are $S^1\\times S^2$, resolving the $S^1\\times S^2$ analogue of the Property R question.","The concordance invariants $\\tau$ and $\\epsilon$ are not zero-trace invariants, assuming the forthcoming identification of the bordered invariants with the usual ones.","For framings in the range $n=0$ or $2-2\\nu(K)