{"id":"17be9a90-b2f4-4f8a-b936-284f8aae7472","arxiv_id":"2607.09026","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The two-view geometric planted matching threshold b=2 is all-or-nothing, while K views enable efficient almost-exact recovery of relative matchings for all b>K/(K-1).","lead":"In high-dimensional geometric planted matching, the known almost-exact recovery threshold b=2 is all-or-nothing: below it, no positive fraction of the matching can be recovered and even Euclidean point-cloud estimation fails. Multiple independent noisy views lower the efficient recovery threshold to b>K/(K-1), so a third view can make recovery possible where two views give nothing.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified","rationale":"The reader correctly isolates the Gaussian high-dimensional regime as the modeling boundary of the work and rates correctness risk low. After a full pass through Appendices A and B I find no load-bearing technical gap inside that regime: the concentration of the free energy, the cycle-type spectral bound, the design of the truncation event E, and the multi-view moderate-deviation + combinatorial gap argument all close. The concrete check proposed above is a routine verification rather than a repair. Consequently the ACCEPT / HIGH-confidence verdict stands unchanged.","tokens_in":38788,"tokens_out":543,"duration_ms":7006,"concrete_test":"Independently re-derive the second-moment ratio bound (A.20) and the truncated first-moment claim of Lemma A.8 for the single numerical value b=1.5 (with the same event E(ε,c0) used in the paper); if the resulting limsup of (1/(n log n)) log E[Z^{2} 1_E] stays ≤ b+2+ζ for arbitrarily small ζ while the first moment remains 1+b/2, the conditional second-moment argument is confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central all-or-nothing claim for two-view geometric matching (Theorem 2.1(i) + Corollary 2.2) and the multi-view algorithmic threshold (Theorem 2.5) rest on standard high-dimensional Gaussian calculations that are carried through carefully. The free-energy upper bound on the restricted Franz–Parisi potential (Proposition A.1) uses chi-squared maxima and a spectral analysis of cycle-type MGFs that correctly identifies the maximizing derangement as a product of 2-cycles; the matching lower bound on the unrestricted free energy (Proposition 2.3 / Theorem A.3) uses a carefully truncated second-moment method whose high-probability event E caps sub-permutation scores at their typical scale, and the Paley–Zygmund step closes the argument for every fixed b<2. The multi-view positive result is a transparent union bound whose worst-case multiplicity pattern is the one-outlier configuration, yielding the clean threshold K/(K-1). No internal inconsistency, missing uniformity, or circular step appears in the appendices. The Gaussian + d=ω(log n) setting is an explicit modeling choice already flagged by the reader; it is not a hidden flaw in the argument as written.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper studies geometric planted matching of n Gaussian points in R^d under the high-dimensional scaling d=ω(log n) and noise variance σ^{2}=d/(b log n). For the classical two-view model it proves that the known almost-exact-recovery threshold b=2 is all-or-nothing: for every fixed b<2 the posterior mass on permutations with overlap ≥δn is at most exp(-C_{b,δ} n log n), which implies that both the permutation MMSE and the Euclidean Y-MMSE of the matched cloud are asymptotically no better than random guessing (Theorem 2.1(i), Corollary 2.2). For the K-view extension the authors give a simple polynomial-time row-wise scoring algorithm that recovers all relative matchings up to o(n) errors whenever b>K/(K-1) (Theorem 2.5). Parallel all-or-nothing statements are obtained for the non-geometric Gaussian weighted matching model.","tokens_in":39012,"tokens_out":816,"duration_ms":7534,"significance":"The work closes a natural gap left by earlier almost-exact-recovery analyses (KNW22, DCK23, WWXY22) by ruling out any nontrivial “something” phase below b=2, both combinatorially and geometrically. The multi-view threshold K/(K-1) is a clean, quantitative demonstration that additional independent views can break the two-view impossibility barrier; already for K=3 the interval 3/2<b<2 becomes efficiently recoverable. The technical core—a carefully truncated second-moment method for the assignment partition function together with an explicit spectral analysis of cycle-type quadratic forms—is self-contained and of independent interest for other high-dimensional matching problems. The arguments are written out in full in the appendices and do not rely on fitted parameters or circular reductions.","major_comments":[],"minor_comments":[{"comment":"The multi-view model is introduced with noise parameter τ while the two-view model uses σ; the asymptotic equivalence σ^{2}∼\tau^{4} is stated only in Remark B.1. A short forward pointer in §2.2 would help the reader compare the two thresholds immediately.","section":null},{"comment":"In the proof overview of Theorem 2.1 the high-probability event E is defined with an auxiliary scale k_{0}=c_{0} n; the subsequent dependence of the second-moment bound on the free parameters ε and c_{0} is correct but slightly opaque. A one-sentence remark that any sufficiently small ε,c_{0} work for b<2 would improve readability.","section":null},{"comment":"The Gaussian weighted matching results (Theorem 2.7, Corollary 2.9) are stated as “analogous” and their proofs are omitted. A brief pointer to the precise places where the geometric argument simplifies (replacement of XZ⊤/√d by an i.i.d. Gaussian matrix) would make the claim fully self-contained.","section":null},{"comment":"Typographical: “Franz–Parisi potential” is introduced without a citation; a standard reference (e.g., Franz–Parisi 1995 or a modern survey) would be appropriate.","section":null}],"recommendation":"accept","confidential_remarks":"The manuscript is technically solid and the all-or-nothing + multi-view message is of clear interest to the high-dimensional statistics / statistical-physics community. I see no reason to delay acceptance; the minor presentation points can be handled at the copy-editing stage."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"This paper closes the remaining gap on the two-view geometric planted matching model in the high-d regime. Prior work already put almost-exact recovery at b>2 and ruled out almost-exact recovery for b<2. What is new is the strong nothing statement: for every fixed b<2 the posterior mass on any positive-overlap set is exp(-C n log n), so both the discrete MMSE and the geometric Y-MMSE are asymptotically no better than the dummy estimators. The multi-view extension is equally clean: a simple row-wise score plus completion recovers all relative matchings almost exactly once b>K/(K-1), so three views already break the two-view barrier for 3/2<b<2.\n\nThe technical work is careful and modular. The negative side uses a truncated second-moment argument on the free energy, with an explicit high-probability event that caps sub-permutation scores at their typical scale, plus a spectral analysis of cycle-type quadratic forms that correctly identifies the worst-case derangement as a product of 2-cycles. The positive side is a transparent union bound whose worst multiplicity pattern is the one-outlier configuration. Both arguments are written out in full; the Gaussian + d=ω(log n) setting is stated up front and is not hidden.\n\nSoft spots are minor and proportional. The entire analysis lives inside the exact Gaussian generative model; non-Gaussian or low-d regimes are left open, as the authors note. The multi-view information-theoretic lower bound is not settled, so we do not yet know whether K/(K-1) is sharp. Neither issue undercuts the theorems that are proved.\n\nThis is for people working on high-dimensional matching, database alignment, or multi-omics integration. The math is solid, the citations are appropriate, and the results are self-contained. I would send it to referees without hesitation and would cite the all-or-nothing and multi-view statements myself.","headline":"Solid all-or-nothing resolution of the two-view geometric matching threshold plus a clean multi-view algorithmic gain already nontrivial at K=3.","tokens_in":39662,"tokens_out":488,"would_cite":true,"duration_ms":5383,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["62H12","60C05","68Q87"],"pacs":[],"model":"grok-4.5","headline":"Below b=2 two-view geometric matching is information-theoretically empty, but a third view recovers all correspondences efficiently.","keywords":["planted matching","geometric matching","all-or-nothing","multi-view recovery","free energy","high-dimensional statistics","posterior concentration"],"falsifier":"Simulate the two-view model at fixed b=1.9 with n=10^4 and d=100 log n; if any estimator recovers more than 5 percent of the matching or reduces Euclidean risk of the matched cloud by more than 5 percent relative to the dummy estimator, the nothing-phase claim fails.","tokens_in":39681,"feed_emoji":"🔍","tokens_out":681,"duration_ms":6219,"temperature":0.7,"pith_summary":"When you observe a high-dimensional point cloud and one noisy, randomly permuted copy of it, earlier work already showed that almost-exact recovery of the hidden matching is possible precisely when a noise parameter b exceeds 2. This paper proves the complementary statement: for every fixed b less than 2 the posterior puts exponentially small mass on any permutation that agrees with the truth on a positive fraction of points. Consequently no estimator can recover a non-vanishing fraction of the matching, and even Euclidean estimation of the matched cloud is asymptotically no better than random guessing. The same paper then introduces a multi-view model in which K independently permuted noisy copies of the same latent cloud are observed. A simple polynomial-time row-wise scoring algorithm recovers every relative matching up to o(n) errors as soon as b exceeds K/(K-1). Already for K=3 the new threshold drops below 2, so a third view can break the two-view impossibility barrier.","feed_headline":"Third view breaks the b=2 matching barrier","feed_subtitle":"Two noisy copies give nothing below b=2; three copies recover everything above 1.5","key_machinery":"Conditional second-moment analysis of the free energy (Franz-Parisi potential) of the posterior partition function, together with a multi-view score that counts consistent pairwise inner products across candidate tuples; the score gap is minimized by one-outlier patterns, which forces the threshold K/(K-1).","core_discovery":"In the high-dimensional Gaussian planted matching model the threshold b=2 is all-or-nothing: for every fixed b<2 the posterior mass on permutations with overlap at least δn decays as exp(-C n log n), so both combinatorial and geometric estimation fail completely. With K independent views the same scaling admits efficient almost-exact recovery of all relative matchings whenever b>K/(K-1).","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["All-or-nothing at b=2; K views recover above K/(K-1)","Third view recovers all matchings past b=1.5 threshold","Multi-view drops planted matching barrier to K/(K-1)","Two views fail below b=2; three views recover everything","High-d matching: third view breaks the impossibility wall"],"cache_read_input_tokens":32896,"weakest_assumption_plain":"Everything rests on exact Gaussian rows for both the latent cloud and every noise matrix, together with the high-dimensional regime d much larger than log n; non-Gaussian or low-dimensional analogues are left open.","fun_headline_variants_meta":{"raw":{"variants":["All-or-nothing at b=2; K views recover above K/(K-1)","Third view recovers all matchings past b=1.5 threshold","Multi-view drops planted matching barrier to K/(K-1)","Two views fail below b=2; three views recover everything","High-d matching: third view breaks the impossibility wall"]},"model":"grok-4.5","effort":"low","cost_usd":0.00391,"raw_usage":{"total_tokens":1173,"prompt_tokens":782,"num_sources_used":0,"completion_tokens":96,"cost_in_usd_ticks":39100000,"prompt_tokens_details":{"text_tokens":782,"audio_tokens":0,"image_tokens":0,"cached_tokens":0},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":295,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":782,"tokens_out":96,"duration_ms":3401,"temperature":1.0,"reasoning_tokens":295,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-13T00:54:05.760952+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Simulate the two-view model at fixed b=1.9 with n=10^4 and d=100 log n; if any estimator recovers more than 5 percent of the matching or reduces Euclidean risk of the matched cloud by more than 5 percent relative to the dummy estimator, the nothing-phase claim fails.","supporting_citations":[],"review_version":1}