{"id":"fac690fb-58ce-44ea-bd5c-889d3cb61b02","arxiv_id":"2605.16733","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Proves sharp operator-norm concentration and expectation bounds for sample cross-covariances of sub-Gaussian and Gaussian vectors, governed by effective ranks of the marginal covariances.","lead":"This paper derives dimension-free concentration bounds on how much a sample cross-covariance matrix deviates from its true mean. A smart generalist might read it for tools that analyze relationships between high-dimensional datasets without the usual dependence on ambient dimension.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"Marginal sub-Gaussianity supplies sub-exponential rather than sub-Gaussian tails on the entries of XY^T","rationale":"The reader correctly flags the sub-Gaussian assumption as the weakest link. The concern above is a precise technical refinement of that assumption: marginal versus joint tails on the cross terms. Because the paper is a pure concentration result whose headline rate hinges on this distinction, verifying the tail class used in the proof is the single most direct way to test whether the claim survives.","tokens_in":1575,"tokens_out":476,"duration_ms":32016,"concrete_test":"Locate the proof of the main high-probability bound (presumably the theorem stated after the abstract). Extract the tail inequality applied to the centered matrices X_i Y_i^T - E[·]. Replace it with the matrix Bernstein inequality for sub-exponential matrices (e.g., Tropp 2015, Thm. 1.3 with Orlicz-1 norm) and recompute the resulting operator-norm deviation; if the bound acquires an extra log(d1+d2) factor or loses the pure effective-rank scaling, the original claim does not follow from marginal sub-Gaussianity alone.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim asserts a high-probability operator-norm bound for the sample cross-covariance that depends only on the effective ranks of the two marginal covariances. For this to hold at the stated rate, the random matrices X_i Y_i^T must satisfy a matrix concentration inequality whose tail is governed by the same effective-rank quantities that appear for covariance estimation. When X and Y are only marginally sub-Gaussian, each entry X_j Y_k is the product of two sub-Gaussian random variables and is therefore sub-exponential (Orlicz norm of order 1). Standard matrix Bernstein or matrix Hoeffding inequalities then produce an extra logarithmic factor in the deviation probability or a worse dependence on the effective ranks; any proof that avoids this degradation must either (a) invoke joint sub-Gaussianity of the pair (X,Y) or (b) employ a more delicate chaining argument that controls the sub-exponential process directly. The manuscript does not appear to state which of these routes is taken, leaving the claimed dimension-free rate dependent on an unverified strengthening of the marginal assumption.","agreement_with_reader":"partial"},"referee_report":{"model":"grok-4.3","summary":"The paper establishes sharp dimension-free concentration and expectation bounds for the deviation of a sample cross-covariance matrix from its mean. For sub-Gaussian random vectors, it proves a high-probability operator-norm bound governed by the effective ranks of the two marginal covariance matrices. In the Gaussian case, it proves a matching expectation lower bound allowing arbitrary correlation between the two random vectors.","tokens_in":1779,"tokens_out":551,"duration_ms":34068,"significance":"If the central claims hold, the results would be significant for high-dimensional statistics: they extend matrix concentration techniques to cross-covariance estimation with rates that depend on effective ranks rather than ambient dimensions, and the Gaussian lower bound holds without restrictions on correlation. This could impact applications in covariance estimation, PCA, and multi-view learning where cross terms appear.","major_comments":[{"comment":"§2, Theorem 2.3 (main high-probability bound): the claimed operator-norm deviation rate depends only on the effective ranks r_X and r_Y under the marginal sub-Gaussian assumption; however, each entry of X_i Y_i^T is a product of two sub-Gaussian variables and hence sub-exponential. Standard matrix Bernstein then introduces an extra log factor or worse rank dependence unless a joint sub-Gaussian assumption or specialized chaining is used. The proof in §4 does not explicitly identify which route is taken, leaving the dimension-free claim load-bearing on an unverified strengthening of the hypothesis.","section":"§2, Theorem 2.3"},{"comment":"§3, Theorem 3.1 (Gaussian expectation lower bound): the matching lower bound is proved only under joint Gaussianity. It is unclear whether the same lower bound holds under the weaker marginal sub-Gaussian assumption used for the upper bound, which would be needed to establish sharpness of the general result.","section":"§3, Theorem 3.1"}],"minor_comments":[{"comment":"Notation for effective ranks r_X and r_Y is introduced in §1 but the precise definition (trace / operator norm or sum of squared eigenvalues) is not restated before the main theorems; a one-line reminder would improve readability.","section":"§1"},{"comment":"The abstract mentions 'sharp' bounds but the introduction does not compare the obtained constants or logarithmic factors to the best known results for ordinary covariance estimation (e.g., Vershynin or Koltchinskii-Lounici). Adding a short comparison paragraph would clarify the improvement.","section":"§1"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the careful reading and valuable comments on our manuscript. Below we respond point by point to the major comments and indicate the revisions we will make.","responses":[{"response":"We appreciate the referee's observation on the technical route taken in the proof. The argument in Section 4 relies on a specialized chaining procedure over nets adapted to the effective-rank subspaces of the marginal covariances, combined with vector sub-Gaussian concentration and a decoupling step that controls the cross term directly. This structure bypasses the standard matrix Bernstein bound on the sub-exponential matrix entries and yields the claimed dimension-free rate. We will add a short explanatory paragraph at the start of Section 4 that outlines this strategy and explicitly contrasts it with a direct application of matrix Bernstein, thereby clarifying the argument under the stated marginal sub-Gaussian hypotheses.","revision_made":"yes","referee_comment":"[§2, Theorem 2.3] §2, Theorem 2.3 (main high-probability bound): the claimed operator-norm deviation rate depends only on the effective ranks r_X and r_Y under the marginal sub-Gaussian assumption; however, each entry of X_i Y_i^T is a product of two sub-Gaussian variables and hence sub-exponential. Standard matrix Bernstein then introduces an extra log factor or worse rank dependence unless a joint sub-Gaussian assumption or specialized chaining is used. The proof in §4 does not explicitly identify which route is taken, leaving the dimension-free claim load-bearing on an unverified strengthening of the hypothesis."},{"response":"The lower bound of Theorem 3.1 is proved under joint Gaussianity because the argument uses the rotational invariance and exact tail behavior available only in that setting; it is designed to demonstrate that the upper-bound rate is optimal when the vectors are jointly Gaussian, even under arbitrary correlation. We do not assert that an identical lower bound holds under the weaker marginal sub-Gaussian assumption, nor does the manuscript claim sharpness of the general upper bound beyond the Gaussian case. We will insert a clarifying remark after Theorem 3.1 and in the introduction stating the scope of the lower bound and noting that extending a matching lower bound to marginal sub-Gaussian vectors is left for future work.","revision_made":"yes","referee_comment":"[§3, Theorem 3.1] §3, Theorem 3.1 (Gaussian expectation lower bound): the matching lower bound is proved only under joint Gaussianity. It is unclear whether the same lower bound holds under the weaker marginal sub-Gaussian assumption used for the upper bound, which would be needed to establish sharpness of the general result."}],"tokens_in":1230,"tokens_out":543,"duration_ms":26427,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The one thing to take away is that this work gives high-probability operator-norm bounds on the sample cross-covariance that depend only on the effective ranks of the two marginal covariances, for sub-Gaussian vectors, and backs it with a matching lower bound on the expectation in the Gaussian case with arbitrary correlation. This is new in the sense that it moves beyond the usual auto-covariance or single-sample settings to handle cross terms while preserving the dimension-free character through effective ranks. The lower bound is particularly useful because it shows the rate is tight even when the vectors are correlated. The paper does a good job laying out the claims clearly and targeting sharpness. If the derivations are clean and avoid unnecessary log factors, it strengthens the case for using these bounds in high-dimensional settings. A potential soft spot is the tail behavior under marginal sub-Gaussianity. The product terms can be sub-exponential, which in standard matrix inequalities often costs an extra log or changes the dependence on rank. The authors must have used either joint sub-Gaussian assumptions or a refined chaining to get the stated rate. The abstract leaves this implicit, so the proofs will need to make the route explicit. This is not a deal-breaker but something a referee should check for tightness. Overall, this is for specialists in high-dimensional probability and statistics who need concentration tools for cross-covariances. Readers working on multivariate estimation or random matrix methods would find the effective-rank bounds practical. The combination of upper and lower bounds makes it worth a serious look from a referee. I recommend putting it through peer review rather than a desk reject.","headline":"This paper gives dimension-free operator-norm bounds for sample cross-covariance deviations controlled by effective ranks, plus a matching Gaussian lower bound even with arbitrary correlation.","tokens_in":2272,"tokens_out":393,"would_cite":false,"duration_ms":46989,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":{"model":"grok-4.3","evidence":[{"relation":"unclear","rs_module":"IndisputableMonolith/Cost/FunctionalEquation","rs_theorem":null,"paper_passage":"Theorem 2.1 ... sub-Gaussian random vectors ... effective ranks r_X, r_Y ... operator-norm bound"}],"headline":"Concentration inequalities for sample cross-covariances lie outside RS forcing chain","alignment":"orthogonal","rationale":"Paper derives operator-norm bounds via quadratic empirical processes, polarization, and Gaussian complexities under marginal sub-Gaussianity; no ratio-symmetric cost J, golden-ratio ladder, 8-tick periodicity, or parameter-free constant derivation appears. Domain (high-dimensional probability) is one on which RS framework has no opinion.","tokens_in":51440,"confidence":"high","tokens_out":177,"duration_ms":10639,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"Sub-Gaussian sample cross-covariances deviate from their mean in operator norm at a rate governed by the effective ranks of the marginal covariances.","keywords":["concentration inequalities","cross-covariance matrix","operator norm","sub-Gaussian random vectors","effective rank","dimension-free bounds","Gaussian lower bounds"],"falsifier":"Generate many samples from a sub-Gaussian distribution with small effective ranks and check whether the observed operator-norm deviation exceeds the bound with probability much larger than the failure probability stated in the theorem.","tokens_in":2454,"feed_emoji":"","tokens_out":587,"duration_ms":43726,"temperature":0.7,"pith_summary":"The paper establishes sharp concentration inequalities for the sample cross-covariance matrix of two random vectors. For sub-Gaussian vectors it derives high-probability bounds on the operator norm deviation that depend only on the effective ranks of the individual covariance matrices. In the special case of Gaussian vectors the bounds are shown to be tight by a matching lower bound on the expected deviation, and this lower bound holds for any level of correlation between the vectors.","feed_headline":"Effective ranks bound sample cross-covariance deviations","feed_subtitle":"High-probability operator-norm bounds for sub-Gaussian vectors match lower bounds for Gaussians with any correlation.","key_machinery":"Effective rank of the marginal covariance matrices, which determines the scaling of the operator-norm concentration bound for the sample cross-covariance.","core_discovery":"This paper establishes sharp dimension-free concentration and expectation bounds for the deviation of a sample cross-covariance matrix from its mean. For sub-Gaussian random vectors, we prove a high-probability operator-norm bound governed by the effective ranks of the two marginal covariance matrices. In the Gaussian case, we prove a matching expectation lower bound, allowing arbitrary correlation between the two random vectors.","pith_inferences":["The same effective-rank technique might apply to other bilinear forms or matrix statistics involving two separate samples.","These bounds could tighten sample-size requirements in applications like canonical correlation analysis or multi-view learning.","Verifying the bounds empirically on synthetic data with controlled effective ranks would test their accuracy."],"forward_implications":["The bounds are dimension-free, so they apply in high-dimensional regimes when effective ranks are moderate.","The results hold with high probability for sub-Gaussian vectors and provide matching lower bounds for Gaussians.","Arbitrary correlation is permitted without worsening the lower bound in the Gaussian setting.","These inequalities provide tools for analyzing statistical procedures that rely on cross-covariance estimates."],"fun_headline_variants":["Effective ranks control cross-covariance deviations","Dimension-free operator-norm bounds for cross-covariances","Sub-Gaussian cross-covariance bounds via effective ranks","Matching lower bounds for Gaussian cross-covariance deviations"],"cache_read_input_tokens":64,"weakest_assumption_plain":"The vectors are assumed to be sub-Gaussian, which ensures the moment and tail conditions used to derive the deviation bounds.","fun_headline_variants_meta":{"raw":{"variants":["Effective ranks control cross-covariance deviations","Dimension-free operator-norm bounds for cross-covariances","Sub-Gaussian cross-covariance bounds via effective ranks","Matching lower bounds for Gaussian cross-covariance deviations"]},"model":"grok-4.3","cost_usd":0.009912,"raw_usage":{"total_tokens":4317,"prompt_tokens":489,"num_sources_used":0,"completion_tokens":57,"cost_in_usd_ticks":99124500,"prompt_tokens_details":{"text_tokens":489,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":3771,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":489,"tokens_out":57,"duration_ms":44259,"temperature":1.0,"reasoning_tokens":3771,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-05-19T20:22:58.620406+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Generate many samples from a sub-Gaussian distribution with small effective ranks and check whether the observed operator-norm deviation exceeds the bound with probability much larger than the failure probability stated in the theorem.","supporting_citations":[],"review_version":1}