- Rating
- 3(+1)
- Confidence
- 3
- Informativeness
- 3
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
The paper learns a population covariance prior and task-specific posteriors from noisy covariance estimates using a known Wishart simulator. Conditional stochastic interpolants implement distribution-valued EM under exact population training and sampling. The paper establishes a sharp identifiability threshold, separates observable convergence from latent recovery, and derives an exact spectral reduction for rotationally invariant priors. Experiments cover synthetic distributions, EEG, and financial returns.
Critical strengths
-
The theory identifies substantive limits of covariance-prior learning. Theorem 1 establishes identifiability exactly when , with indistinguishable smooth, compactly supported, rotationally invariant priors below the threshold. Theorem 2 carefully separates observation-space convergence, weak latent recovery under compact spectral bounds, and stronger recovery requiring inverse stability. [pp. 3–4; Appendices B–D]
-
The spectral reduction preserves full posterior uncertainty. The conditional orientation law is independent of the unknown spectral prior, so exact completion adds no KL error. The accompanying analysis distinguishes population eigenvalues, empirical-direction variances, and loss-specific Bayes estimates, avoiding the incorrect identification of aligned eigenvalue samples with matrix posterior samples. [Theorem 4, p. 6; Appendix G]
Critical weaknesses
-
The likelihood and oracle specifications are inconsistent. Finance uses and forecast degrees of freedom , but evaluates covariance targets formed from 21 trading days. Every simulated forecast has rank at most six, whereas those targets are generally full rank. Consequently, the specified singular predictive law cannot provide the claimed full-matrix predictive density. A covariance-dependent expression evaluated outside that support would be a surrogate score rather than the stated predictive density. Directional calibration and realized portfolio returns are not automatically invalidated. [Appendices J.1/J.4–J.5, pp. 42, 45–47; Figure 4c; Table 6] Separately, the main AR(1) model and quadrature oracle use lognormal decay , while Appendix J.2 specifies uniform correlation . These priors are not equivalent under , leaving the largest reported oracle comparison unresolved. [pp. 7, 43]
-
The main empirical claims exceed the stronger validation. Main-text EEG accuracies are 0.852 overall and 0.956 on the confident half; five-fold subject-disjoint evaluation gives 0.806 and 0.911, and the appendix explicitly states that the larger single-split gains do not carry over. [p. 8; Appendix J.3/Table 3, pp. 44–45] The principal empirical conclusions should reflect the subject-disjoint evaluation and distinguish it from the favorable single split.
Questions for authors
-
What probability law produced the financial matrix log-scores at ? Please provide a valid-support reevaluation, or identify and justify the surrogate score while separating its conclusions from portfolio returns.
-
Which AR(1) prior generated Table 1 and its oracle? Please reconcile the generator and quadrature specification and establish whether corrected definitions leave the reported comparison unchanged.
-
Can the main EEG comparison be stated using the five-fold subject-disjoint results, including baseline differences and variability? Resolving the score/oracle definitions and showing which gains survive these checks would support a rating of 3.
5 comments
revised PDF
The revised PDF is uploaded, and the finance proper scores are in the general comment above. Scoring the 21 forward daily returns, SC-SI gains +1.21 [0.97, 1.52] nats per day over the sample covariance under a Gaussian predictive and +0.48 [0.45, 0.52] under a Student- predictive, and it is best under both. The paired gains over OAS (+0.27 [0.25, 0.29]) and IW-ML (+0.18 [0.16, 0.20]) are positive in every volatility tercile and every test decade. The AR(1) specification is corrected in §4.1 and Appendix J.2, and the five-fold EEG results are in §4.2 and Tables 2–3. We hope this settles the three points you raised.
Dear Reviewer, as the discussion period closes on Nov 18, we wanted to check whether our reply and the revised PDF address your concerns. In particular, the three points you listed (finance score, AR(1) oracle, EEG evaluation) are now resolved in the revision, with the proper finance scores in our general comment. We would be glad to clarify anything further before then. Thank you again for your review.
We thank the reviewer for their careful reading. All three issues are correct, and we fixed them as follows.
Q1. Finance log-score with Nf=6<d=12. The reviewer is right. Appendix A defines the channel density for N<d relative to a reference measure, but that density lives on matrices of rank at most Nf. The realized 21-day matrices have full rank, so they fall outside its support. Since terms free of C were dropped, what we computed for each posterior draw Cj is −2Nf[logdetCj+tr(Cj−1Cf)]=21Nf∑t=121logN(rt;0,Cj)+const, a tempered Gaussian log-likelihood of the 21 forward daily returns rt, with temperature s=Nf/21. It is a surrogate score, not the stated predictive density.
In the revision we name it as such and add two proper scores of the 21 returns (results in a follow-up):
Neither score enters the portfolio. Coverage of forward directional variances is unaffected: given C, v⊤Cfv is (v⊤Cv)χNf2/Nf.
We also clarify the role of N and Nf. They are effective degrees of freedom, chosen together on a validation fold that precedes all test folds (Table 4). They absorb heavy tails and the drift between the estimation and forecast windows. A new result gives this a precise meaning: if non-Gaussian samples are treated as Gaussian, the iteration converges to the prior under which the Wishart channel best explains the observed covariances, and Theorem 2 holds relative to that prior. Choosing N is then model selection between well-defined projections. Fig. 4c is labeled as a tempered likelihood and shown next to the proper scores (Fig. 4d).
Q2. AR(1) prior. The reviewer is right that the two descriptions are inconsistent. Table 1 and its quadrature oracle both use Cij=ρ∣i−j∣ with ρ∼U(0.3,0.9). The lognormal-γ description (§4.1, J.2) referred to a variant not used for Table 1 and is removed; the bound 1.0 in J.2 was a typo for 0.9. We also corrected the stated sample size: all synthetic runs use N=40, not the N=28 written in the submission.
In the rerun (longer training, fresh observations at every iteration), SC-SI is 0.03 nats from the oracle on AR(1), with every W1 (median over tasks) at most 0.38 posterior s.d. IW-ML is 9.7 nats below the oracle (23.3 in the submission, whose log-scores used only J=256 draws; we now use J=2⋅106). The conclusion of the submission is unchanged: an inverse-Wishart prior cannot follow this family, and SC-SI can.
Q3. EEG, five-fold evaluation. Agreed. The main text now reports five-fold subject-disjoint cross-validation over all 109 subjects, and the 29-subject split moves to the appendix. Its test subjects turned out to be easier for every method: the sample-covariance accuracy was 0.820 there, against 0.790 over all subjects.
Log-score and calibration, where the main evidence lies:
Classification, pooled over the five folds (3052 windows):
The main-text claim becomes: a clear gain in log-score and calibration, and a modest but significant gain in classification accuracy and NLL.
Summary. You asked us to resolve the score and oracle definitions and to show which gains survive. The AR(1) oracle is consistent and the comparison stands. The finance score is correctly named, with proper scores to follow, and the portfolio result does not depend on it. On all 109 EEG subjects the log-score and calibration gains survive, and the classification gain is smaller but significant. We hope this meets the conditions you set for a positive rating.