{"id":"cf2c28f0-43a9-4fb8-9dc2-97f96dd80e19","arxiv_id":"2608.12183","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"In Gaussian multi-index models, the paper rigorously characterizes the bulk spectrum and BBP-type phase transition of matrix-valued spectral estimators, and proves the AMP-derived preprocessing map attains the minimal spectral threshold, resolving Conjecture 3.10 of Defilippis et al. (2025).","lead":"The paper proves a sharp sample-size phase transition for a family of spectral estimators that recover hidden low-dimensional structure from high-dimensional nonlinear data, and identifies the preprocessing rule with the provably lowest sample requirement. It settles a 2025 conjecture by showing that simple eigenvector-based methods can match message-passing algorithms without needing any informative initialization.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Optimality theorem is conditional on Assumption A.4; for channels with conditionally singular C(y) the AMP-derived T* is inadmissible, so the unqualified AMP-coincidence claim is not established.","rationale":"I read the proof of Theorem 1.10, especially Section 7, carefully. The universal lower bound (Lemma 7.1) is internally consistent: the reduction from H_α(θ)w = 0 to the eigenvalue equation for E[A ⊗ (C-I)], the similarity argument with the stability operator K_θ, Lemma 7.3 using Corollary 3.17, and the Kronecker Cauchy-Schwarz inequality of Lemma 7.4 all check out. The upper bound for T* (Lemmas 7.6 and 7.7) is also coherent: the rescaled MDE fixed point, the strict stability argument via R - K[R] ≻ 0, and the singular Schur complement at x = 1/α are mutually consistent. Thus I found no internal logical error in the main theorem under its stated assumptions. The single load-bearing concern is Assumption A.4: it is exactly what makes T* admissible, and without it the AMP-derived map may be unbounded or undefined. The paper is transparent about this, and the reader's verdict correctly identified it as the weakest assumption. Since Theorem 1.10 is precisely scoped to A.4, the appropriate verdict is unchanged, but the abstract's unqualified phrasing overstates the generality. The proposed check on the noiseless single-index model would settle whether A.4 is merely a technical convenience or an essential restriction on the claimed AMP-spectral coincidence.","tokens_in":75238,"tokens_out":38076,"duration_ms":344132,"concrete_test":"Analyze the noiseless single-index channel r = 1, P(y|s) = δ(y - s), for which C(y) = s^2 and A.4 fails. First compute the AMP weak-recovery threshold from the state evolution of [63] (with denoiser f(y) = E[s|y] = y) and compare it with 1/α*_c = E[(s^2-1)^2] = 2. If the AMP threshold is below 1/2, the unqualified coincidence claim fails outside A.4. Second, define bounded truncations T_L(y) = 1 - 1/min(max(s^2, L^{-1}), L) and compute α_{c,min}(T_L) numerically via the scalar MDE; if α_{c,min}(T_L) does not converge to 1/2 as L → ∞, then A.4 is not a removable technical condition and the optimality theorem genuinely depends on it.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing premise is Assumption A.4 (C(y) ⪰ cI almost surely, with P(C(y) ≠ I_r) > 0). It is needed to make T*(y) = I_{p*} - C_*(y)^{-1} from Definition 1.8 well-defined and uniformly bounded, and hence an admissible preprocessing map in the class over which optimality is claimed in Theorem 1.10. The universal lower bound in Lemma 7.1 does not use A.4, but the attaining map T* does. For natural multi-index channels, A.4 can fail: for example, in the noiseless single-index model y = s_1 (r = 1), C(y) = s_1^2, which is positive but has no uniform lower bound. Then C(y)^{-1} is unbounded and T* is not an admissible bounded map, so Theorem 1.10 says nothing about the optimal spectral threshold in that channel. The abstract, however, states the coincidence with the AMP weak-recovery threshold without this qualification. Thus the headline claim 'its transition coincides with the AMP weak-recovery threshold, proving the general spectral conjecture' is only proved under A.4; whether the restriction is technical or essential is left open. The internal proof under A.4 appears coherent; the soft spot is the scope of the claimed optimality.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a random matrix theory for matrix-valued spectral estimators of the form D_n = (1/n) Σ_i T(y_i) ⊗ x_i x_i^T in Gaussian multi-index models, where T is a bounded symmetric matrix-valued preprocessing map of fixed dimension p. The main results are: (i) almost-sure convergence of the empirical spectral measure of D_n to a deterministic compactly supported law characterized by a matrix Dyson equation with a nonlinear self-energy; (ii) a BBP-type phase transition for the largest eigenvalue, with an explicit deterministic outlier equation and a sharp threshold in terms of the function h_α(x); (iii) weak recovery of the latent subspace from the leading eigenvector in the supercritical regime; and (iv) an optimality theorem stating that the AMP-derived preprocessing T_*(y) = I_{p*} - C_*(y)^{-1} is optimal among all bounded matrix-valued preprocessing maps of any fixed dimension, with threshold 1/α*_c = ||E_y[(C(y)-I_r)⊗(C(y)-I_r)]||_op, coinciding with the AMP weak-recovery threshold. The proofs are detailed and self-contained, with Sections 3–7 providing the technical arguments, and Appendices A–C giving heuristic motivation, a variational edge characterization, and auxiliary results.","tokens_in":75249,"tokens_out":8514,"duration_ms":83405,"significance":"If the main claims hold, this is a substantial contribution to high-dimensional statistics and random matrix theory. It resolves a natural conjecture about the reach of spectral methods in multi-index models, removes the simultaneous-diagonalizability assumption of earlier work, and establishes a rigorous optimality result for the AMP-derived preprocessing. The technical toolbox is impressive: a nonlinear matrix Dyson equation, a self-adjoint linearization for sign-indefinite covariance models, a spectral comparison theorem to a free Gaussian model, a Fock-space variational bound for the free edge, and a finite-dimensional reduction of the outlier problem. The proofs are structured, detailed, and appear internally consistent. The paper also honestly discloses the limitation that Proposition 1.7 does not characterize the intermediate regime [α_c,min, α_c,max] for general maps. The main weakness is the scope of the optimality theorem, which is conditional on Assumption A.4 and whose abstract formulation overstates the unqualified validity.","major_comments":[{"comment":"The optimality claim is conditional on Assumption A.4, which requires C(y) ⪰ c I_r almost surely. This condition fails for natural channels such as the noiseless single-index model y = s_1 with r=1, where C(y) = s_1^2 is positive but not uniformly bounded below. In such a channel the proposed optimal preprocessing T_*(y) = 1 - C(y)^{-1} is unbounded and therefore not admissible in the class of bounded preprocessing maps over which optimality is claimed. Consequently Theorem 1.10 does not establish the AMP-threshold coincidence or the general spectral conjecture for these channels. The paper does not discuss whether A.4 is technical or essential; in particular, it does not analyze whether a truncated or regularized version of T_* attains the same threshold α*_c. Since this is a restriction on the channel, not on the estimator, it is load-bearing for the headline claim. I ask the authors to qualify the statement of Theorem 1.10 and the abstract, and to add a discussion (or a rigorous approximation argument) addressing channels where C(y) is singular or only positive semidefinite with non-uniform lower bound.","section":"Section 1.3, Theorem 1.10 and Assumption A.4"},{"comment":"The abstract states that the AMP-derived preprocessing is optimal among all bounded matrix-valued preprocessing maps and that its transition coincides with the AMP weak-recovery threshold, 'proving the general spectral conjecture of [Defilippis et al., 2025]'. As written, this suggests a fully general result. However, the proof in Section 7 requires Assumption A.4 for the definition and admissibility of T_*; outside that assumption the theorem does not apply. Moreover, the statement 'Theorem 1.10 proves Conjecture 3.10 of [26]' needs clarification: if the conjecture was originally stated for all Gaussian multi-index channels, the present paper proves a restricted version. The authors should explicitly state the precise class of channels for which the conjecture is proved and, if the original conjecture was broader, indicate the open case. This is more than a presentation issue because it affects the advertised scope of the main optimality result.","section":"Abstract and Introduction (claims about the general spectral conjecture)"}],"minor_comments":[{"comment":"The abstract should mention Assumption A.4 explicitly, or rephrase the optimality claim to 'under the nondegeneracy condition A.4', so that the conditional nature of the result is visible to the reader.","section":"Abstract"},{"comment":"The first paragraph contains a typo: 'two independent aprts' should be 'two independent parts'.","section":"Section 6.3"},{"comment":"There is a spurious period in the sentence 'We next show that a separating outlier. exists'; it should read 'a separating outlier exists'.","section":"Section 7.2"},{"comment":"The symbol ⊗ is used both for the Kronecker product and for the tensor product in the Fock-space construction. While the meaning is clear from context, a brief remark in the notation section would improve readability.","section":"Section 1.1 and Section 5.2"}],"recommendation":"major_revision","confidential_remarks":"This is a strong, proof-heavy paper with a significant technical contribution. The main issue is the discrepancy between the unqualified abstract claim and the actual scope of Theorem 1.10, which requires Assumption A.4. I believe this is addressable by adding a careful discussion of the scope, qualifying the abstract, and ideally providing an approximation argument for channels failing A.4. If the authors can do that, the paper would be suitable for acceptance. I recommend major revision to allow for this clarification."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here's the take: this paper does what the abstract promises—mostly. It removes the simultaneous-diagonalizability restriction that blocked prior work and develops a genuine noncommutative theory for matrix-weighted covariance operators. The bulk MDE with nonlinear self-energy, the upper-edge confinement via free Gaussian comparison, and the BBP transition are all new and look technically sound. The proof of optimality for the AMP-derived preprocessing T* is the centerpiece, and it's convincing: the threshold alpha*_c is derived from the channel's second-order fluctuations, not fitted, and the lower bound over all bounded preprocessing maps is a real theorem, not a heuristic.\n\nThe soft spot is exactly the one the stress-test flagged. Assumption A.4—uniform positive lower bound C(y) ⪰ cI—is load-bearing. It makes T*(y) = I - C(y)^{-1} well-defined and bounded, hence admissible. For natural channels like noiseless single-index with y = s, C(y)=s^2 and A.4 fails; T* is unbounded and outside the admissible class. The theorem then says nothing about whether a spectral method can reach the AMP threshold in that channel. The abstract's unqualified sentence claiming coincidence with the AMP threshold overstates what is proved. The paper itself states A.4 explicitly before Theorem 1.10, so the theorem is scoped honestly; the abstract is where the overreach sits. Whether the assumption is technical or essential is left open—that's a fair criticism but it doesn't undermine the proofs under A.4.\n\nI spot-checked the leave-one-out decomposition, the sign-symmetry argument for the off-diagonal term, and the stability analysis of the MDE. Everything I checked is consistent. I did not verify every line of the deep edge-confinement argument, which leans on Bandeira-Cipolloni-Schröder-van Handel and a Lehner-type Fock-space bound; those are the parts a referee should read most carefully. No circularity, no fitted parameters.\n\nThis is a paper for specialists in random matrix theory and high-dimensional statistics, and it deserves a serious referee. The restriction A.4 should be discussed honestly in any revision, and the abstract should carry the qualification. But the core contribution is substantial and the proofs are in scope. Send it to review.","headline":"Serious, proof-heavy paper that substantially advances the spectral theory of Gaussian multi-index models; the optimality theorem is real but its scope is narrower than the abstract claims.","tokens_in":76021,"tokens_out":2461,"would_cite":true,"duration_ms":24719,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["60B20","62H12","68Q87"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that, in Gaussian multi-index models, the AMP-derived preprocessing map $T_*(y)=I_{p*}-C_*(y)^{-1}$ attains the optimal spectral weak-recovery threshold among all bounded matrix-valued preprocessing maps of any fixed…","keywords":["Gaussian multi-index models","matrix-valued spectral methods","matrix-weighted covariance matrices","BBP transitions","weak subspace recovery","approximate message passing","matrix-valued self-consistent equations","spectral phase transitions"],"falsifier":"Construct a two-index channel in which $C(y)$ is singular or arbitrarily close to singular on a set of positive probability, meaning that a bounded approximation of $T_*$ is the only admissible option; if the spectral transition of that approximation occurs strictly below $\\alpha^*_c$, or if the formula $1/\\alpha^*_c=\\|\\mathbb{E}[(C(y)-I_r)^{\\otimes 2}]\\|_{\\mathrm{op}}$ fails to predict the empirical outlier location, the optimality theorem would be refuted. A direct check in a non-commuting two-index model would also settle whether the predicted threshold matches the AMP threshold numerically.","tokens_in":74799,"feed_emoji":"📊","tokens_out":5555,"duration_ms":56305,"temperature":0.7,"pith_summary":"The paper studies spectral estimators for recovering a hidden $r$-dimensional subspace from nonlinear Gaussian observations, in the proportional regime where sample size and dimension grow at the same rate. It proves a complete spectral phase transition for the matrix-weighted covariance estimator $D_n=\\frac{1}{n}\\sum_i T(y_i)\\otimes x_i x_i^\\top$: below a deterministic threshold the largest eigenvalue stays at the bulk edge, above it an outlier separates and the associated eigenvector achieves weak recovery. Its central result is that the AMP-derived preprocessing map $T_*(y)=I_{p*}-C_*(y)^{-1}$ is optimal among all bounded symmetric matrix-valued preprocessing maps of any fixed dimension, with threshold $1/\\alpha^*_c=\\|\\mathbb{E}_y[(C(y)-I_r)\\otimes (C(y)-I_r)]\\|_{\\mathrm{op}}$ exactly matching the AMP weak-recovery threshold. If correct, this settles the general spectral conjecture that no fixed-dimensional bounded spectral preprocessing can beat the AMP threshold in these models.","feed_headline":"Spectral estimators hit the AMP recovery limit","feed_subtitle":"The AMP-derived preprocessing map is proven optimal among all bounded fixed-dimension spectral methods.","key_machinery":"The engine is the matrix-weighted covariance matrix $D_n=\\frac{1}{n}\\sum_{i=1}^n T(y_i)\\otimes x_i x_i^\\top$ and the matrix Dyson equation $M_\\alpha(z)^{-1}=-z I_p+\\mathbb{E}_y\\big[T(y)(I_p+\\alpha^{-1}M_\\alpha(z)T(y))^{-1}\\big]$, which characterizes the deterministic bulk spectrum. The spectral transition is governed by the deterministic outlier function $h_\\alpha(x)=\\lambda_1\\big(\\mathbb{E}_y[T(y)(I_p+\\alpha^{-1}M_\\alpha(x)T(y))^{-1}\\otimes C(y)]\\big)-x$; the edge gap $\\Delta_T(\\alpha)=\\lim_{x\\downarrow \\lambda_+(\\alpha)}h_\\alpha(x)$ decides whether an outlier separates. Optimality is carried by the Perron reduction of the operator $A[H]=\\mathbb{E}_y[(C(y)-I_r)H(C(y)-I_r)]$, which produces the preprocessing map $T_*(y)=I_{p*}-C_*(y)^{-1}$ whose threshold is exactly $\\alpha^*_c$.","core_discovery":"The paper establishes a sharp, universal lower bound on the spectral weak-recovery threshold: for every fixed $p\\ge 1$ and every bounded measurable preprocessing map $T$, one has $\\alpha_{c,\\min}(T)\\ge \\alpha^*_c$, where $1/\\alpha^*_c=\\|\\mathbb{E}_y[(C(y)-I_r)\\otimes(C(y)-I_r)]\\|_{\\mathrm{op}}$. Equality is attained by the preprocessing map $T_*(y)=I_{p*}-C_*(y)^{-1}$, constructed from the Perron eigenmatrix $H_*$ of the operator $A[H]=\\mathbb{E}_y[(C(y)-I_r)H(C(y)-I_r)]$; for this map the lower and upper critical sampling ratios coincide, $\\alpha_{c,\\min}(T_*)=\\alpha_{c,\\max}(T_*)=\\alpha^*_c$. Combined with the phase-transition theorem, this proves that the AMP weak-recovery threshold of Troiani et al. is attainable by a spectral method without side information, and that no bounded fixed-dimensional matrix-valued preprocessing can do better, thereby proving Conjecture 3.10 of Defilippis et al.","pith_inferences":["The optimality is proved inside the class of bounded, fixed-dimension preprocessing maps; it does not by itself rule out spectral estimators built from unbounded maps or from features whose dimension grows with $n$, so the 'no spectral method beats AMP' reading should be restricted to this class.","The threshold formula $1/\\alpha^*_c=\\|\\mathbb{E}[(C(y)-I_r)^{\\otimes 2}]\\|_{\\mathrm{op}}$ identifies the size of channel fluctuations as the sole resource for spectral recovery; a testable consequence is that any channel with $C(y)=I_r$ almost surely is spectrally unrecoverable at every finite $\\alpha$.","The Perron-reduction construction suggests a general recipe: linearize the AMP update around its uninformative fixed point and form the resulting matrix-weighted covariance; the spectral transition will coincide with the AMP instability point whenever the conditional second-moment operator has a positive Perron eigenmatrix.","A natural extension not pursued in the paper is to replace Gaussian covariates by an orthogonally invariant design and check whether the same formula, with $C(y)$ and the design covariance, still predicts the transition."],"forward_implications":["Below the threshold $\\alpha^*_c$, the largest eigenvalue of $D_n$ sticks to the bulk edge, so no spectral estimator of this class has a nonvanishing overlap with the hidden subspace.","Above $\\alpha^*_c$, an outlier separates at a location given by the unique zero of $h_\\alpha$, and every leading eigenvector achieves weak recovery of the latent subspace with overlap bounded away from zero.","The AMP weak-recovery threshold can be attained by a spectral method without side information or an informative initialization.","No choice of working dimension $p$ or bounded matrix-valued preprocessing rule can improve on $\\alpha^*_c$, so the threshold depends only on the second-order fluctuations of the conditional second moment $C(y)$."],"supporting_citations":[{"why":"Introduces the AMP-derived matrix-valued spectral estimator and states the general spectral conjecture that this paper proves.","marker":"[26]"},{"why":"Establishes the sharp AMP weak-recovery threshold in Gaussian multi-index models, the benchmark the spectral method must match.","marker":"[63]"},{"why":"Solves the single-index case with optimal scalar preprocessing, the $r=1$ benchmark that the matrix-valued theory generalizes.","marker":"[46]"},{"why":"Provides the spectral comparison theorem used, conditionally on the responses, to confine the upper edge of the noise spectrum.","marker":"[10]"},{"why":"Supplies the Fock-space variational argument adapted to bound the upper edge of the associated free Gaussian operator.","marker":"[38]"},{"why":"Gives the fixed-point and positivity arguments used to establish existence and uniqueness of the solution to the matrix Dyson equation.","marker":"[34]"},{"why":"Optimizes scalar-preprocessed spectral estimators in multi-index models under simultaneous diagonalizability, the restrictive setting removed here.","marker":"[37]"}],"fun_headline_variants":["Spectral estimator hits AMP weak-recovery limit","Spectral method ties AMP weak-recovery threshold","Optimal spectral preprocessing attains AMP recovery","Without side info: spectral reaches AMP optimum"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is Assumption A.4: the conditional second moment $C(y)=\\mathbb{E}[ss^\\top|y]$ is bounded below by a positive constant times the identity almost surely and is not identically the identity; if some latent direction is determined essentially exactly, the map $T_*(y)=I_{p*}-C_*(y)^{-1}$ is not bounded and the optimality claim does not apply.","fun_headline_variants_meta":{"raw":{"variants":["Spectral estimator hits AMP weak-recovery limit","Spectral method ties AMP weak-recovery threshold","Optimal spectral preprocessing attains AMP recovery","Without side info: spectral reaches AMP optimum"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001177,"raw_usage":{"total_tokens":4958,"prompt_tokens":1135,"completion_tokens":3823,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":751,"completion_tokens_details":{"reasoning_tokens":3766}},"tokens_in":751,"tokens_out":3823,"duration_ms":29323,"temperature":1.0,"reasoning_tokens":3766,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:14:48.886816+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a two-index channel in which $C(y)$ is singular or arbitrarily close to singular on a set of positive probability, meaning that a bounded approximation of $T_*$ is the only admissible option; if the spectral transition of that approximation occurs strictly below $\\alpha^*_c$, or if the formula $1/\\alpha^*_c=\\|\\mathbb{E}[(C(y)-I_r)^{\\otimes 2}]\\|_{\\mathrm{op}}$ fails to predict the empirical outlier location, the optimality theorem would be refuted. A direct check in a non-commuting two-index model would also settle whether the predicted threshold matches the AMP threshold numerically.","supporting_citations":[{"cited_title":"Optimal spectral transitions in high-dimensional multi-index models","cited_arxiv_id":null,"evidence_quote":"Introduces the AMP-derived matrix-valued spectral estimator and states the general spectral conjecture that this paper proves."},{"cited_title":"Fundamental computational limits of weak learnability in high-dimensional multi-index models","cited_arxiv_id":null,"evidence_quote":"Establishes the sharp AMP weak-recovery threshold in Gaussian multi-index models, the benchmark the spectral method must match."},{"cited_title":"Mondelli and A","cited_arxiv_id":null,"evidence_quote":"Solves the single-index case with optimal scalar preprocessing, the $r=1$ benchmark that the matrix-valued theory generalizes."},{"cited_title":"Bandeira, G","cited_arxiv_id":null,"evidence_quote":"Provides the spectral comparison theorem used, conditionally on the responses, to confine the upper edge of the noise spectrum."},{"cited_title":"Lehner,Computing norms of free operators with matrix coefficients, Amer","cited_arxiv_id":null,"evidence_quote":"Supplies the Fock-space variational argument adapted to bound the upper edge of the associated free Gaussian operator."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the fixed-point and positivity arguments used to establish existence and uniqueness of the solution to the matrix Dyson equation."}],"review_version":1}