{"id":"afe737e5-8fe8-4f44-9cec-77a8a64e6279","arxiv_id":"2607.20760","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A coskewness-maximizing split objective in a twoblock regression tree recovers piecewise linear regimes better than covariance splits and matches black-box ensembles on two multivariate benchmarks.","lead":"Twoblock clustering trees are new regression trees that split data by maximizing the skewness between predictors and outcomes, then fit a local, dimension-reduced linear model inside each group. A smart generalist should read this if they want an interpretable model that, on two public benchmarks, comes close to or beats random forests and neural networks while using a fraction of the parameters.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Table 1 confounds objective and cut-search: ARI gains credited to coskewness may stem from the k-means (t1,u1) split, not the third-moment objective; the reader flagged this, and the paper's own Figure 3 shows the k-means variant also lifts covariance ARI.","rationale":"The reader's weakest_assumption focuses on the unverified mixture-skewness mechanism (Section 5) and the local-convergence of HOPM (Section 3.3). My concern overlaps but is more specific and more directly falsifiable: the headline numbers in Table 1 confound the objective with the cut-search heuristic. The paper says 'the coskewness objective wins on ARI at every noise level' but the winning config pairs coskewness with k-means, while the covariance baseline uses exhaustive variance cut. Figure 3 shows a k-means lift for the covariance objective, so the comparison is not apples-to-apples. This is not a claim of dishonesty—the paper is open about the flag and shows Figure 3—but the central claim's attribution is not established by the reported evidence. I agree with the reader's 'CONDITIONAL' verdict. The concern lands because it directly targets the load-bearing assertion; a single re-analysis can settle it. If the re-analysis shows coskewness still wins under the exhaustive cut, my concern is neutralized and the paper's claim stands. I do not REJECT because the paper is generally honest, includes limitations (small-n failure, coupled-variant inferiority, local convergence), and the real-world results are suggestive even if not conclusive. The concrete test is computationally feasible (reuse existing code, five seeds, six factorial cells), so CONDITIONAL is appropriate: acceptance should require this unconfounded comparison.","tokens_in":14896,"tokens_out":2135,"duration_ms":16259,"concrete_test":"Re-run the four-regime simulation of Section 6.1 (all six rows of Table 1) under the full factorial of objective (covariance, coskewness, combined) × cut-search (exhaustive variance on t1, k-means (t1,u1)) with the same five seeds. Report the ARI for the coskewness objective with the exhaustive variance cut; if this value is at or near the covariance-with-exhaustive value (≈0.44–0.47) rather than near the coskewness-with-k-means value (0.52–0.64), the Table 1 attribution to coskewness fails and the headline gain is driven by the k-means cut search. Additionally, compute the marginal ARI gain of k-means for the covariance objective; if this gain is comparable to the gain credited to coskewness, the claim that 'the coskewness objective recovers regimes markedly better' needs to be rephrased to 'the k-means (t1,u1) partition recovers regimes markedly better when combined with a higher-momen","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central empirical claim in Sections 6.2 and 7.2 is that the coskewness objective recovers piecewise-linear regimes markedly better than the covariance objective that underlies PLS-Trees (Table 1: ARI 0.52–0.64 vs 0.40–0.47), and that this regime recovery is the mechanism behind the decoupled tree's predictive competitiveness. But Table 1 changes two factors at once: the objective (covariance vs coskewness) AND the cut-search (exhaustive variance cut on t1 vs k-means partition in the (t1,u1) plane). Figure 3 shows that the k-means variant lifts ARI for all objectives, including covariance, meaning the headline margin partly belongs to the k-means search, not to coskewness. The paper never reports the coskewness objective under the exhaustive variance cut in Table 1, so the reader cannot isolate the coskewness contribution. Section 5.2 identifies the k-means partition as 'selected by a separate flag' that 'pairs particularly well with the coskewness objective', but no interaction table is given. This is a confounded comparison in the paper's central evidence. The reader's weakest_assumption ('the mixture-skewness mechanism... never verified analytically') captures a related but distinct issue; the confound is more directly the load-bearing weakness because it undermines the clean attribution of Table 1. The decoupled tree and real-world results do not resolve this: Table 3 compares coskewness-split vs combined-split trees, and the black-box baselines are undocumented in tuning, but the core mechanistic claim relies on Table 1, which is unconfounded only if the k-means interaction is negligible. Since the paper itself provides a hint that k-means helps covariance (Figure 3 dashed lines rise for cov), the concern is concrete and testable.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes twoblock clustering trees (tbtree), a deterministic regression tree for multivariate responses in which each node fits a twoblock dimension-reduction model that defines the split direction, and each leaf carries a local multivariate linear twoblock model. A new coskewness-based twoblock decomposition is introduced, solved by a matrix-free higher-order power method. The paper argues through a controlled simulation that the coskewness objective recovers planted non-normal structure that covariance is blind to, and through a four-regime simulation that it recovers piecewise-linear regimes better than the covariance objective underlying PLS-Trees. The default configuration decouples the split objective (coskewness or combined) from the leaf objective (covariance), and two real-world benchmarks (energy efficiency, gas-turbine emissions) are used to claim that this decoupled tree matches or beats black-box ensembles while remaining a readable single tree.","tokens_in":15254,"tokens_out":5542,"duration_ms":51889,"significance":"If the empirical claims hold, the contribution is significant: a single, interpretable regression tree with local linear models approaching ensemble-level predictive accuracy, and the first twoblock dimension reduction based on coskewness, would be a useful addition to the interpretable-ML toolbox. The paper is honest about limitations (the small-n failure, the coupled variant's inferiority, and the predictive cost of coskewness splits are all disclosed) and ships reproducible notebooks. The matrix-free solver avoids materializing O(p^2q) tensors, which is a genuine algorithmic strength. However, the central attribution of regime-recovery gains to the coskewness objective is currently confounded with the cut-search heuristic, and the supporting simulation evidence lacks error bars. These issues are fixable with additional analysis and experiments, so the manuscript merits a major revision rather than rejection.","major_comments":[{"comment":"The headline comparison changes two factors simultaneously: the objective (covariance vs coskewness) and the cut-search (exhaustive variance cut on t1 vs k-means partition in the (t1,u1) plane). Figure 3 shows that the k-means variant also lifts ARI for the covariance objective, so the margin in Table 1 cannot be cleanly attributed to the third-moment objective. To support the claim that 'the coskewness objective recovers the regimes markedly better than the covariance objective,' report the full factorial design (all objectives × both cut-search heuristics), or at least coskewness+exhaustive and covariance+k-means, and quantify the interaction. Without this, the attribution is not established.","section":"Section 6.2, Table 1"},{"comment":"The simulation conclusions rest on five seeds with no standard errors, confidence intervals, or per-seed values. Some of the reported ARI differences (e.g., 0.52 vs 0.47 at sigma_eps=0.1 in Scenario A) are small relative to what seed-to-seed variability could plausibly be. Report variability over seeds (or increase the number of seeds) for all cells of the design, and make the per-seed results available. This is load-bearing because the main claim is a comparative one based on these margins.","section":"Sections 6.1–6.2"},{"comment":"The combined objective is a convex combination of E[tu] and E[t^2u]. These two terms have different units and can differ by orders of magnitude even after autoscaling, so gamma is not a scale-invariant knob across datasets. The sharp basin transition in Figure 1(b) may be specific to that design. Specify how the two moments are normalized in the combined objective (e.g., dividing each term by a norm or a reference scale), or justify explicitly why autoscaling alone makes gamma interpretable. This affects the interpretation of all 'combined' results in Sections 6–7.","section":"Section 3.2, Eq. (5)"},{"comment":"The k-means partition in the (t1,u1) score plane is central to the claimed ARI advantage in Table 1, yet the manuscript only says it is 'catalogued in the Supplementary Material' and does not describe the algorithm, initialization, or how its boundary is snapped to a threshold on t1. No supplementary material is provided with the manuscript. This is missing support for a load-bearing component. Provide the algorithmic details, or remove the reliance on an unavailable reference.","section":"Section 5.2, k-means partition"},{"comment":"The black-box baselines (TCN, Cubist, random forest, MLP, CART) are reported without hyperparameter settings, tuning protocol, or repeated evaluation, and the temporal split is a single realization. The claim that tbtree 'matches or beats every black-box baseline on scale-weighted error' needs either a clear statement of default configurations and a sensitivity analysis, or a more complete benchmark protocol. At minimum, specify the software versions, the random seeds, and the hyperparameters for each baseline.","section":"Section 7.2, Table 3"}],"minor_comments":[{"comment":"Several typos and misspellings need correction: 'peicewise' (Abstract), 'thrid' (Section 3), 'twolock clusterin' and 'decison' (Section 5.1), 'bne explained' (Section 3.1), 'prtoposed' (Section 1), 'of of' (Section 2).","section":"Throughout"},{"comment":"The text says the combined-split tree has eleven leaves, while the figure caption says '10 leaves' and appears to show ten leaves. Verify the count and make the text and caption consistent.","section":"Section 7.1, Figure 4"},{"comment":"The sentence 'because the k-means partition ... is selected by a separate flag; it turns out to pair particularly well with the coskewness objective' is vague. If the k-means variant is recommended, state how it is chosen in practice and whether it is used in the real-data benchmarks.","section":"Section 5.2"},{"comment":"The software section says the package 'will be released to the public in the upcoming release.' For reproducibility, please provide a versioned release (e.g., DOI or tag) or indicate that the examples/notebooks can be run with the current development version.","section":"Section 9"}],"recommendation":"major_revision","confidential_remarks":"The central empirical claim is currently confounded, and the missing supplementary description of the k-means split and the absence of error bars are fixable but substantive. I recommend asking for a factorial analysis of the simulation and a description of the k-means heuristic before further consideration. The paper's honest reporting of negative results is commendable and should be preserved."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I read this one carefully because the novelty is real. The coskewness-based twoblock decomposition — maximizing the third-order cross-moment of the score pair instead of the covariance — is a genuine addition to the dimension-reduction literature, and the matrix-free HOPM that never forms the moment tensor is a real algorithmic improvement. Nesting that in a deterministic regression tree with decoupled split and leaf objectives is a thoughtful design: third-moment routing for regime separation, covariance leaves for predictive accuracy.\n\nThe best part of the paper is the Section 4 isolation study. The antithetic sampling zeroes out every odd sample moment of z exactly, so cov(z,y)=0 by construction while E[z²y]=2. Showing that the covariance objective cannot rank the planted direction above the decoy, and that coskewness loads on it at |w|≈0.99, is transparent and honest. The paper also discloses its own failures — the predictive cost of skewness-aligned splits, the small-n concrete-slump degradation, the local convergence of the solver, the coupled-variant inferiority. That all earns credit.\n\nThe soft spots sit in the empirical core, and the stress-test note has the right one. Table 1 changes two factors at once: the covariance objective is paired with the exhaustive variance cut, while the coskewness objective is paired with the k-means (t1,u1) split. Figure 3 shows the k-means search lifts ARI for the covariance objective too, and the text concedes k-means 'pairs particularly well' with coskewness. So part of the headline margin plausibly belongs to the cut search, not the third-moment objective. Since regime recovery is the central claim, the paper needs the full objective×cut-search interaction table before the attribution is clean. The fix is easy — report coskewness under the exhaustive cut — but as presented, the comparison is confounded.\n\nThe rest is the usual statistical laundry: five seeds and no error bars anywhere, single train/test splits on both real benchmarks, undocumented baseline tuning, and code that is promised but not yet public. The mixture-skewness mechanism is asserted rather than derived, and a balanced four-regime mixture can have low overall skewness; the paper never checks that the third-order moments actually dominate at the tested noise levels. All fixable, all worth fixing.\n\nI agree with the CONDITIONAL verdict. No load-bearing error or circularity; the claims are just under-supported as presented. The paper is worth a serious referee — send it out, and require the interaction table, error bars or repeated-split variance, baseline disclosure, and public code.","headline":"A genuinely new coskewness twoblock estimator and a clean isolation study, but the regime-recovery headline rests on a confounded Table 1 and five seeds with no error bars.","tokens_in":15937,"tokens_out":5139,"would_cite":true,"duration_ms":45173,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["62H30","62H25","62J05","62-08"],"pacs":[],"model":"deepseek-v4-flash","headline":"A regression tree that splits on the coskewness of predictor–response scores recovers the piecewise-linear regimes that covariance-based splits miss; decoupling the split from the leaf model lets it match black-box regressors.","keywords":["twoblock clustering tree","coskewness","third-order moments","higher-order power method","regression trees","multivariate response","regime recovery","piecewise linear models"],"falsifier":"Run the Section 6 four-regime design with regimes mirrored so the mixture has zero skewness but nonzero covariance structure; if a coskewness-split tree still beats a covariance-split tree on adjusted Rand index, the gain is not coming from third moments and the central mechanism is falsified.","tokens_in":14688,"feed_emoji":"🌳","tokens_out":7221,"duration_ms":63010,"temperature":0.7,"pith_summary":"The paper introduces the twoblock clustering tree (tbtree), a deterministic decision tree for multivariate responses that places a local multivariate linear model in each leaf. Its central claim is that regime boundaries in piecewise-linear data are third-moment phenomena: a node spanning two regimes is a non-Gaussian mixture, and a split objective that maximizes the coskewness of the predictor and response scores separates those regimes more cleanly than the usual covariance objective. Because third moments are built into the split but not into prediction, the paper recommends a decoupled tree — a coskewness or combined split, with ordinary covariance models refit in the leaves — and shows it recovers simulated four-regime structure with higher adjusted Rand index while using fewer leaves. On two real benchmarks the decoupled tree matches or beats a 200-tree random forest and a temporal convolutional network on scale-weighted error while remaining a single tree of 10–16 readable local models.","feed_headline":"Split on coskewness to recover regimes covariance trees miss","feed_subtitle":"A single interpretable tree with third-moment splits matches random-forest accuracy on two benchmarks.","key_machinery":"The load-bearing mechanism is the coskewness twoblock decomposition, a joint low-dimensional reduction of predictor and response blocks: instead of maximizing the covariance E[tu] between predictor score t and response score u, the latent directions solve max E[t²u] (predictor side) and max E[tu²] (response side), or a convex combination controlled by γ. The paper solves this with a matrix-free, block-symmetric higher-order power method that never materializes the third-moment tensor, so each direction costs O(n(p+q)) and is initialized deterministically from the covariance warm start plus coordinate starts. The tree consumes only the first X-score t1=Xw1 as a scalar ordering, searches a cut","core_discovery":"On its own terms, the paper establishes that the first predictor score of a twoblock model fitted to maximize coskewness — rather than covariance — is a better ordering for cutting a regression tree when the response surface is piecewise linear across several regimes. In a four-regime simulation, the coskewness objective with a k-means cut in the (t1,u1) score plane reaches an adjusted Rand index of 0.52–0.64 against 0.40–0.47 for the covariance objective, and the advantage widens as noise increases, while producing about six leaves for four regimes instead of ten to twelve. The paper then shows that the same tree, with leaves refit under the covariance objective, matches or beats every blac","pith_inferences":["An independent attribution test not run in the paper: compare a covariance-split tree using the k-means (t1,u1) cut against a coskewness-split tree using the exhaustive variance cut. The paper reports best configurations, so part of the ARI gain could come from the cut heuristic rather than the objective.","The balanced four-regime design (n_k=100 per regime) is roughly symmetric in its mixture structure, where odd moments can partially cancel. An asymmetric or unequal-size version would test whether coskewness remains the regime signal or whether the reported gain is specific to the design.","The solver's reliance on deterministic multi-start means the extracted direction depends on warm starts; a sensitivity analysis that perturbs the starts would show how stable the split directions are, especially on wide data.","If the mechanism is right, a natural bridge not explored in the paper is using coskewness-based twoblock scores for unsupervised regime discovery when the response block is replaced by auxiliary variables."],"forward_implications":["For piecewise multivariate linear data, splitting on coskewness finds the true regimes more reliably than covariance splits, and the advantage grows as noise increases (ARI 0.61 vs 0.40 at the highest noise level in the slope-only scenario).","A single decoupled tbtree can reach ensemble-level accuracy: best scale-weighted error on gas-turbine emissions among all compared methods, and energy-efficiency R² within noise of a 200-tree random forest.","The resulting model is a shallow tree of 10–16 leaf models whose coefficients can be read to see which predictor drives the response in each operating regime — something black-box ensembles cannot provide.","Decoupling matters: using coskewness in the leaves hurts prediction; using it only to choose the split preserves regime recovery while keeping covariance-based leaf regressions accurate.","The combined objective with γ≈0.5 is the recommended compromise on smooth nonlinear data; pure coskewness is best on sharply piecewise data, and covariance remains the default on small samples."],"fun_headline_variants":["Coskewness splits reveal piecewise linear regimes covariance misses","Covariance fails, coskewness succeeds: new regression tree","Third-moment splits boost regime recovery in twoblock trees","Match random forests with interpretable coskewness tree"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that a node containing several regimes has third-order joint moments that separate the regimes more strongly than covariances do, and that the local power method finds that maximizing direction; the paper states this as an assumption and never verifies the moment dominance analytically for its simulation design.","fun_headline_variants_meta":{"raw":{"variants":["Coskewness splits reveal piecewise linear regimes covariance misses","Covariance fails, coskewness succeeds: new regression tree","Third-moment splits boost regime recovery in twoblock trees","Match random forests with interpretable coskewness tree"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000942,"raw_usage":{"total_tokens":3855,"prompt_tokens":731,"completion_tokens":3124,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":475,"completion_tokens_details":{"reasoning_tokens":3052}},"tokens_in":475,"tokens_out":3124,"duration_ms":21131,"temperature":1.0,"reasoning_tokens":3052,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T09:28:38.063662+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the Section 6 four-regime design with regimes mirrored so the mixture has zero skewness but nonzero covariance structure; if a coskewness-split tree still beats a covariance-split tree on adjusted Rand index, the gain is not coming from third moments and the central mechanism is falsified.","supporting_citations":[],"review_version":1}