{"id":"ec128990-256d-4e39-bf13-5f6549d9251e","arxiv_id":"2607.26906","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"CC-geodesic Kakeya sets in the first Heisenberg group have sharp Heisenberg Hausdorff dimension 3 (even when compact), not the predicted 4.","lead":"The natural Kakeya conjecture in the Heisenberg group fails for Carnot–Carathéodory geodesics: such sets need only Heisenberg dimension 3, not 4, and compact examples attain this bound. The paper also builds full-dimension zero-measure examples and one-dimensional fixed-curvature near-misses.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified","rationale":"The reader correctly isolates the sole external citation as the weakest link, yet that link is not load-bearing in the sense of correctness risk: the comparison theorems of [3] are classical and apply directly to the sets constructed here. All other steps—parametrizations (2.2)–(2.3), the explicit translates of Lemma 3.1, the inductive length control in the proof of Theorem 1.2, and the elementary Lipschitz estimates—are self-contained and verifiable by direct differentiation and the group law. The abstract typo (“4” versus “3”) is already flagged by the reader and does not affect the body. Consequently the ACCEPT verdict and low correctness-risk assessment stand.","tokens_in":13381,"tokens_out":539,"duration_ms":211588,"concrete_test":"Re-derive the three coordinate limits of F_j(κ,s) as κ→0 by Taylor expansion (or symbolic L’Hôpital) up to order 2 and confirm that the resulting map is C¹ on a neighbourhood of the rectangle [0,ρ_j]×[0,2]; if the first derivatives fail to match the claimed horizontal segment (0,s,0), the compactness-plus-dimension argument for E_c would need repair.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central sharpness claims (Theorems 1.1–1.2) rest on explicit left-translates of all γ_θ,k into Σ ∪ Π (Lemma 3.1 plus the horizontal plane) and into the compact union E_c = D² ∪ ⋃ E_j (via the screw-motion identity (4.3) and the length check u+1 ≤ ℓ_j(k)). Both constructions are elementary and close under the group law. The only external input is the dimension comparison that C¹ Euclidean surfaces (and Euclidean-Lipschitz images of compact 2-dimensional parameter domains) have Heisenberg Hausdorff dimension exactly 3 (cited from Balogh–Durand-Cartagena–Fässler–Mattila–Tyson [3, p. 414 and Thm. 2.7]). That comparison is a standard, repeatedly used fact in sub-Riemannian GMT and applies verbatim to the smooth embedded surface Σ, to Π, and to each F_j(A_j). No internal gap, hidden non-minimality, or failure of the length/cut-time constraints appears.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper studies CC-geodesic Kakeya sets in the first Heisenberg group H¹: Borel sets containing a left translate of every unit-speed minimising CC-geodesic segment of length 1 issuing from the identity. The natural analogue of the Kakeya conjecture would require Heisenberg Hausdorff dimension 4. The authors prove that every such set has dim_H^{H¹} ≥ 3 (inherited from the horizontal case K_CC ⊂ K({0})), and that 3 is sharp, both in general (Theorem 1.1, via the explicit union Σ ∪ Π of two smooth Euclidean surfaces) and among compact sets (Theorem 1.2, via a compact packing E_c = D² ∪ ⋃_j E_j of geodesic arcs). They also produce a CC-geodesic Kakeya set of full dimension 4 and Lebesgue measure zero (Corollary 1.3), and, for each fixed curvature κ ∈ (0, 2π], a compact curvature-κ Kakeya set of Euclidean and Heisenberg Hausdorff dimension 1 (Theorem 1.4).","tokens_in":13542,"tokens_out":963,"duration_ms":35696,"significance":"The work answers Lukyanenko’s open question and shows that the natural full-dimension prediction fails: nonzero-curvature geodesics do not raise the sharp lower bound above the horizontal value 3. Sharpness is obtained by fully explicit left-translation constructions (Lemma 3.1 and the screw-motion identity (4.3)) whose dimension upper bounds rest only on standard comparison theorems for C¹ surfaces and Lipschitz images in H¹. The fixed-curvature near-miss at dimension 1 is a clean and striking counterpart. The arguments are elementary, self-contained, and free of circularity; the contribution is solid and appropriate for a serious journal in geometric measure theory / harmonic analysis.","major_comments":[{"comment":"Abstract vs. Theorems 1.1–1.2: the abstract states that “their sharp lower bound is 4”, while the body (Introduction, Theorem 1.1, Theorem 1.2, and the constructions) consistently proves that the sharp lower bound is 3 and that the prediction of 4 fails. This is a load-bearing misstatement of the main result and must be corrected to “3” before publication; as written, the abstract contradicts the theorems.","section":"Abstract"}],"minor_comments":[{"comment":"In the abstract, insert the missing space in “is4” (and likewise check “dimension4”, “measure0”, etc., for consistent spacing).","section":"Abstract"},{"comment":"Page 1 / Introduction: the sentence “A fact that at first glance supports this conjecture is that, if one uses only horizontal unit segments… then the sharp lower bound… is 3” is slightly awkward; a brief clarifying clause that horizontal segments are the k = 0 members of the family would help non-specialist readers.","section":"Section 1"},{"comment":"Lemma 3.1: the two cases k > 0 and k < 0 differ only by a constant phase shift of π in the t-coordinate; a one-line remark that the same computation covers both signs (or a unified formula) would shorten the write-up.","section":"Lemma 3.1"},{"comment":"Proof of Theorem 1.2: the removable-singularity claim for F_j at κ = 0 is correct by Taylor expansion, but citing the explicit limiting expressions (or noting that the coordinate functions extend C^∞) would make the Lipschitz bound immediate without appealing to a neighbourhood of the rectangle.","section":"Section 4.3"},{"comment":"References: the arXiv number and year of the Wang–Zahl Kakeya paper [26] should be double-checked against the final published version if available at proof stage.","section":"References"}],"recommendation":"minor_revision","confidential_remarks":"The sole substantive defect is the abstract’s erroneous claim that the sharp lower bound is 4 rather than 3—an obvious transcription slip that does not affect the correctness of the body. Once corrected, the paper is ready for acceptance; I see no reason for further delay. Fit with math.CA / geometric measure theory venues is excellent."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The one thing worth knowing is that the natural CC-analogue of the Kakeya conjecture in the first Heisenberg group fails: requiring left translates of every unit-speed CC-geodesic (not just the horizontal ones) does not push the sharp Heisenberg Hausdorff dimension above 3. That bound stays sharp for compact sets as well. The paper also gives a full-dimension-4 zero-measure example and, for each fixed nonzero curvature, a compact near-miss of dimension 1.\n\nWhat is new is the move from the horizontal class K({0}) treated by Liu and by Fässler–Pinamonti–Wald to the full family K_CC. The lower bound of 3 is immediate from the inclusion, so the real work is the sharpness constructions. Lemma 3.1 puts every nonzero-curvature segment into the smooth surface Σ by an explicit left translate; union with the horizontal plane gives the unbounded example of dimension 3. The compact set E_c is built by packing truncated screw-motion arcs into countably many Lipschitz images of compact 2-dimensional parameter domains plus a disk; the length and cut-time checks close cleanly under the group law. The fixed-curvature sets are just single extended trajectories, which is the right observation once curvature is locked.\n\nThe soft spots are minor and proportional. The dimension comparison that C¹ Euclidean surfaces (and Euclidean-Lipschitz images of 2-dimensional domains) have Heisenberg dimension exactly 3 is cited from Balogh et al.; it is standard in the subfield and applies verbatim to the sets written down here. There is a typo in one version of the abstract claiming the lower bound is 4, but the body and theorems correctly state 3. No circularity, no free parameters, no hidden non-minimality.\n\nThis is for people already working on Kakeya-type problems in Carnot groups or sub-Riemannian GMT. The arguments are coordinate-level and short; a serious referee will finish them in an afternoon. I would send it to peer review without hesitation and would bring it to reading group if we have anyone in the area.","headline":"Clean negative answer to Lukyanenko’s question: full CC-geodesic Kakeya sets in H¹ still only force Heisenberg dimension 3, even when compact, and the constructions are elementary and checkable.","tokens_in":14262,"tokens_out":539,"would_cite":true,"duration_ms":15224,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["28A78","53C17","22E25"],"pacs":[],"model":"grok-4.5","headline":"CC-geodesic Kakeya sets in the first Heisenberg group have sharp Heisenberg dimension 3, not the expected 4.","keywords":["Kakeya sets","CC-geodesics","Heisenberg group","Hausdorff dimension","Carnot–Carathéodory geometry","curvature-κ Kakeya sets","sub-Riemannian geometry"],"falsifier":"Exhibit a CC-geodesic Kakeya set whose Heisenberg Hausdorff dimension is strictly less than 3, or show that the constructed surface Σ fails to contain a left translate of some nonzero-curvature unit geodesic.","tokens_in":14174,"feed_emoji":"📐","tokens_out":922,"duration_ms":28518,"temperature":0.7,"pith_summary":"In the first Heisenberg group, a natural Kakeya-type set is one that contains a left translate of every unit-speed Carnot–Carathéodory geodesic segment leaving the identity. The homogeneous dimension is 4, so the direct analogue of the classical Kakeya conjecture would demand that every such set has Heisenberg Hausdorff dimension 4. This paper shows that prediction is false: every CC-geodesic Kakeya set has dimension at least 3, and 3 is sharp—even when the set is required to be compact. The same lower bound was already known for the weaker horizontal-only problem; the new work shows that forcing in all nonzero-curvature geodesics does not raise the bound. The paper also builds a (non-compact) example of full dimension 4 with zero Lebesgue measure, and shows that if one freezes a single positive curvature then compact examples of dimension 1 exist.","feed_headline":"Heisenberg Kakeya sets need only dimension 3, not 4","feed_subtitle":"Forcing every CC-geodesic into the set does not raise the sharp lower bound above the horizontal case","key_machinery":"The surface Σ parametrized by Φ(r,φ)=(r cos φ, r sin φ, r²φ/2), together with the horizontal plane Π. Suitable left translates of every nonzero-curvature unit geodesic land inside Σ, while the zero-curvature segments sit in Π; their union is therefore a CC-geodesic Kakeya set of Heisenberg dimension 3.","core_discovery":"Every CC-geodesic Kakeya set in the first Heisenberg group has Heisenberg Hausdorff dimension at least 3, and this lower bound is attained by an explicit Borel set (and even by a compact one). Consequently the natural full-dimension-4 conjecture fails. Separately, for each fixed curvature κ in (0,2π] there exist compact curvature-κ Kakeya sets of Euclidean and Heisenberg dimension exactly 1.","pith_inferences":["The fixed-curvature dimension-1 examples suggest that any attempt to restore a dimension-4 conjecture would need to quantify over a positive-measure set of curvatures, not a single κ.","The same surface-plus-plane construction may adapt to higher-step or higher-dimensional Heisenberg groups once the geodesic equations are known explicitly.","Because the sharpness example is a union of two Euclidean surfaces, quantitative Kakeya maximal inequalities in H¹ cannot improve beyond the horizontal case without new geometric input."],"forward_implications":["The CC-geodesic Kakeya conjecture in H¹ is false at the level of Hausdorff dimension; the sharp lower bound is 3 rather than 4.","Sharpness holds already inside the compact class, so unboundedness is not needed to reach dimension 3.","Full dimension 4 is still compatible with zero Lebesgue measure for CC-geodesic Kakeya sets.","Fixing a single positive curvature collapses the problem dramatically: compact examples of dimension 1 exist.","Horizontal-only Kakeya lower bounds already control the full geodesic family; extra curvature does not improve the constant."],"fun_headline_variants":["CC-geodesic Kakeya sets in Heisenberg group hit dimension 3 sharply","Heisenberg CC-Kakeya bound is 3, not the expected 4","Full dimension-4 conjecture fails for CC-geodesic Kakeya sets","Compact CC-geodesic Kakeya sets attain Heisenberg dimension 3","Curvature-κ Kakeya sets can have Euclidean and Heisenberg dim 1"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The dimension comparison that every smooth two-dimensional Euclidean surface inside the Heisenberg group has Heisenberg Hausdorff dimension exactly 3.","fun_headline_variants_meta":{"raw":{"variants":["CC-geodesic Kakeya sets in Heisenberg group hit dimension 3 sharply","Heisenberg CC-Kakeya bound is 3, not the expected 4","Full dimension-4 conjecture fails for CC-geodesic Kakeya sets","Compact CC-geodesic Kakeya sets attain Heisenberg dimension 3","Curvature-κ Kakeya sets can have Euclidean and Heisenberg dim 1"]},"model":"grok-4.5","effort":"low","cost_usd":0.004668,"raw_usage":{"total_tokens":1281,"prompt_tokens":700,"num_sources_used":0,"completion_tokens":82,"cost_in_usd_ticks":46684000,"prompt_tokens_details":{"text_tokens":700,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":499,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":700,"tokens_out":82,"duration_ms":9144,"temperature":1.0,"reasoning_tokens":499,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-30T17:27:02.575354+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Exhibit a CC-geodesic Kakeya set whose Heisenberg Hausdorff dimension is strictly less than 3, or show that the constructed surface Σ fails to contain a left translate of some nonzero-curvature unit geodesic.","supporting_citations":[],"review_version":1}