Nonlocality Effect in the Alpha decay half-lives of superheavy nuclei with XGBRegressor

Chen Wu; Jinyu Hu

REVIEW 3 major objections 2 minor 2 references

Nonlocality Effect in the Alpha decay half-lives of superheavy nuclei with XGBRegressor

T0 review · 3 major / 2 minor · reviewed 2026-07-13 · grok-4.5

Challenge this review Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read An improved two-potential model of alpha-nucleus nonlocality, tuned with gradient boosting, cuts half-life RMS error by 74.8% on 599 nuclei and predicts Z=117-120 superheavy decays in line with DZR and MUDL.

desk verdict We only have the abstract for this nuclear TPA/XGB paper; the cached “full text” is Moretti’s unrelated QFT localization paper, so the 74.8% RMS claim cannot be audited. read the letter →

arxiv 2604.04187 v1 submitted 2026-04-05 nucl-th

Jinyu Hu , Chen Wu This is my paper · ORCID

classification nucl-th

keywords alphadecayhalf-livessuperheavynucleitwo-potentialapproachnonlocalityXGBRegressorodd-Aodd-odd

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper generalizes the alpha-nucleus nonlocality effect, previously treated mainly for even-even systems, to odd-A and odd-odd nuclei inside the two-potential approach. Coordinate-dependent parameters that encode that nonlocality are optimized with an XGBRegressor gradient-boosting model. On a set of 599 nuclei with Z from 52 to 118 the improved scheme lowers the root-mean-square deviation of calculated alpha-decay half-lives by 74.8 percent relative to the original TPA. The same calibrated model is then used, together with the independent DZR and MUDL formulas, to forecast half-lives of 142 superheavy candidates with Z=117-120; the three sets of predictions agree closely, and the improved TPA results are nearly identical to those of DZR. The practical claim is that a carefully tuned nonlocal correction can both describe known alpha emitters far more accurately and supply mutually consistent forecasts for the heaviest nuclei still under experimental study.

What carries the argument

The improved two-potential approach (TPA) that embeds a generalized, coordinate-dependent nonlocality effect for all parity classes and whose free parameters are fitted by XGBRegressor.

What would settle it

A blinded or hold-out evaluation of the improved TPA on a reserved subset of the 599 known half-lives, or new measured half-lives for Z=117-120 nuclei that systematically deviate from the joint TPA-DZR-MUDL predictions.

Watch

Extended reading notes

Core claim

Generalizing alpha-nucleus nonlocality to odd-A and odd-odd systems inside the two-potential approach, and optimizing the resulting coordinate-dependent parameters with gradient boosting, produces an improved TPA whose half-life RMS deviation on 599 nuclei (Z=52-118) is 74.8 percent lower than that of the original TPA, while its predictions for 142 superheavy nuclei (Z=117-120) stay in close agreement with the DZR and MUDL models.

Load-bearing premise

That the 74.8 percent RMS reduction reflects genuine physical improvement rather than an in-sample fit of flexible parameters on the same nuclei used to quote the metric, without a clear hold-out or validation protocol.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The one thing you need to know: I cannot review the actual manuscript of Hu & Wu (arXiv:2604.04187). What was supplied as full text is Moretti’s math-ph paper on relativistic localization in QFT (2604.04173). Everything below is abstract-only plus that mismatch.

From the abstract alone, the contribution is incremental but real in its niche: they extend Medeiros-style alpha–nucleus nonlocality inside the two-potential approach to odd-A and odd-odd systems, fit the new coordinate-dependent nonlocality parameters with XGBRegressor, report a 74.8% RMS drop versus original TPA on 599 nuclei (Z=52–118), and then give half-life predictions for 142 superheavy candidates (Z=117–120) that sit close to DZR and MUDL (nearly identical to DZR). That is a legitimate applied-nuclear extension of an existing program, not a new framework.

The soft spot is load-bearing and not minor. Flexible, coordinate-dependent parameters plus a gradient-boosting regressor can absorb residuals on the same set used to quote RMS. The abstract never states train/validation/test splits, cross-validation, early stopping, feature list, or whether the 599-nucleus RMS is out-of-sample. Without that, the 74.8% figure does not establish physical improvement; it may be in-sample fit quality. Multi-model agreement on Z=117–120 is useful as a consistency check but does not fix the evaluation design.

Who cares: people who already work alpha half-lives, TPA, and superheavy search planning. Nobody outside that strip needs this until the protocol is clear.

I would not bring it to reading group until the real PDF (equations, residual plots, splits, code) is in hand. I would not cite it on present evidence. If a complete nuclear manuscript matching the abstract exists, a serious editor should still send it to referees—the program is standard enough and the claim is sharp enough to deserve a methods check—not desk-reject on abstract alone. Right now the central number is simply unauditable.

Referee Report

3 major / 2 minor

Summary. The submission is presented as a nuclear-theory paper that generalizes the alpha–nucleus nonlocality effect in the two-potential approach (TPA) to odd-A and odd-odd nuclei, optimizes coordinate-dependent nonlocality parameters with XGBRegressor, reports a 74.8% reduction in RMS half-life deviation relative to the original TPA on 599 nuclei (Z=52–118), and predicts half-lives for 142 superheavy nuclei (Z=117–120) in close agreement with the DZR and MUDL models. The supplied full manuscript text, however, is an unrelated mathematical-physics article (Moretti, arXiv:2604.04173) on relativistic spatial localization observables constructed from the stress–energy tensor in free QFT. None of the nuclear-physics claims, methods, or numerical results can be checked against the body of the document.

Significance. If the abstract’s claims were supported by a matching, carefully validated manuscript, an improved TPA with a large RMS gain and consistent superheavy predictions would be of genuine interest for alpha-decay systematics and SHN half-life estimates. As submitted, that significance cannot be assessed: the body of the paper does not contain TPA, nonlocality parameters, XGBRegressor, the 599-nucleus sample, or the Z=117–120 predictions. The only load-bearing issue is the complete mismatch between title/abstract and full text.

major comments (3)

Title, abstract, and arXiv identifier (2604.04187, nucl-th) describe a nuclear TPA/XGB study of alpha-decay half-lives, but the full manuscript is Moretti’s QFT localization paper (arXiv:2604.04173, math-ph). There is no TPA formalism, no nonlocality parameters, no XGBRegressor protocol, no 599-nucleus dataset, and no superheavy predictions. The central claims cannot be audited.
Even taking the abstract alone, the headline 74.8% RMS reduction is not accompanied by any train/validation/test split, cross-validation, regularization, or hold-out protocol for the XGB-optimized coordinate-dependent parameters. Without that protocol, the improvement may be an in-sample fit rather than a physical gain; the missing nuclear manuscript prevents any check of residual plots, feature set, or out-of-sample performance.
Agreement of improved TPA with DZR and MUDL for Z=117–120 is offered as support for the predictions, but no tables, uncertainties, or model definitions appear in the supplied text. Reliability of the 142-nucleus forecast cannot be evaluated.

minor comments (2)

The abstract cites Medeiros [1], a previous study [2], DZR [3], and MUDL [4] without bibliographic detail in the provided materials; a correct nuclear manuscript would need a complete reference list and clear definitions of those models.
If a corrected nuclear manuscript is resubmitted, it should state explicitly whether the 599 nuclei used for the RMS figure are disjoint from the set used to train XGBRegressor, and should report hyperparameters and feature construction.

Circularity Check

2 steps flagged · score 6.0 of 10

Abstract optimizes coordinate-dependent nonlocality parameters with XGBRegressor then quotes 74.8% RMS improvement on the same 599 nuclei without any stated train/hold-out protocol—classic fitted-input-as-metric circularity; superheavy Z=117–120 forecasts are only partly insulated.

fitted input called prediction [Abstract (RMS claim on 599 nuclei)]
"The coordinate-dependent parameters introduced by this nonlocality are optimized using an advanced gradient boosting regression model. This improved TPA is applied to calculate the α-decay half-lives of 599 nuclei with Z = 52-118, yielding a root-mean-square (RMS) deviation that is 74.8% lower than that of the original TPA."

Parameters of the improved model are optimized by XGBRegressor, then the same improved model is scored by RMS on the 599 nuclei (Z=52–118). The abstract never states that this set is held out, cross-validated, or otherwise independent of the fit. With flexible coordinate-dependent nonlocality parameters plus gradient boosting, the 74.8% RMS reduction is statistically forced in-sample and is not shown to be an out-of-sample physical improvement.
self citation load bearing [Abstract (opening premise)]
"Building on the work of E. L. Medeiros [1] and our previous study [2], we generalize the alpha-nucleus nonlocality effect to odd-A and odd-odd nuclei within the two-potential approach (TPA) framework."

The nonlocality framework and its prior calibration are imported from the authors’ own previous study [2] as the starting point for the present generalization and XGB re-optimization. This is secondary to the fit/metric circularity above and is normal cumulative work, but it means the coordinate-dependent nonlocality ansatz is not independently re-derived here—only extended and re-fit.

full rationale

Only the abstract of arXiv:2604.04187 is available for the claimed nuclear TPA/XGB paper; the supplied full manuscript is a different work (Moretti, arXiv:2604.04173 on QFT localization), so equations, tables, and protocol cannot be audited beyond the abstract. From the abstract alone the load-bearing chain is: (i) introduce coordinate-dependent nonlocality parameters generalizing Medeiros and the authors’ prior study; (ii) optimize those parameters with XGBRegressor; (iii) report RMS on α-decay half-lives of 599 nuclei (Z=52–118) as 74.8% lower than original TPA. No train/validation/test split, cross-validation, early stopping, or regularization is stated, so the headline metric is consistent with in-sample residual absorption by a flexible gradient-boosting fit of coordinate-dependent parameters—i.e., the improvement can reduce to the fit by construction. Predictions for 142 superheavy nuclei (Z=117–120) and agreement with DZR/MUDL are a separate, less circular step if those nuclei were held out of the XGB fit, but they do not salvage the central 74.8% claim. Self-citation of the authors’ previous study [2] is normal scaffolding, not load-bearing uniqueness. Score 6 reflects partial circularity on the main quantitative claim, not total definitional collapse of the whole paper.

Assumptions & free parameters 2 free parameters · 4 assumptions · 1 invented entities

Abstract-only ledger for a phenomenological nuclear model. Load-bearing inputs are the TPA framework, Medeiros-style nonlocality, extension to odd systems, and ML-tuned coordinate-dependent parameters. Free parameters are the optimized nonlocality coefficients (count and values unknown from abstract). No new particles; the ‘invented’ objects are the coordinate-dependent nonlocality parameters and the XGB-tuned map from nuclear properties to those parameters.

free parameters (2)

coordinate-dependent nonlocality parameters (TPA)
Abstract states these are introduced by nonlocality and optimized with XGBRegressor; values, count, and priors are not given in the abstract but the 74.8% RMS claim depends on them.
XGBRegressor hyperparameters and feature set
Gradient boosting model used to optimize the nonlocality parameters; unspecified in the abstract yet controls the fit quality.

assumptions (4)

domain assumption Two-potential approach (TPA) is an adequate base model for alpha-decay half-lives across Z=52–118.
Central calculations and the baseline ‘original TPA’ comparison rest on this standard nuclear-theory framework.
domain assumption Alpha-nucleus nonlocality (Medeiros-type) can be generalized from even systems to odd-A and odd-odd nuclei via coordinate-dependent parameters.
Stated as the paper’s generalization step; physical validity for odd systems is assumed rather than derived from first principles in the abstract.
domain assumption RMS deviation on the 599-nucleus set is a valid figure of merit for model quality and for comparing to original TPA.
Headline 74.8% improvement is defined entirely through this metric.
ad hoc to paper Agreement among improved TPA, DZR, and MUDL supports reliability of Z=117–120 predictions.
Cross-model concordance is used as validation; models may share systematics, so agreement is not independent experimental confirmation.

invented entities (1)

Coordinate-dependent nonlocality parameters for odd-A and odd-odd nuclei in TPA
purpose: Extend nonlocality corrections beyond even-even systems and provide knobs for XGB optimization of half-lives.
Abstract presents these as introduced by the nonlocality generalization; no independent experimental handle beyond improved half-life fits is stated.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Nonlocality Effect in the Alpha decay half-lives of superheavy nuclei with XGBRegressor." pith.science (2026). https://pith.science/paper/2604.04187

@misc{pith2026260404187,
  author       = {Pith},
  title        = {Pith review of: Nonlocality Effect in the Alpha decay half-lives of superheavy nuclei with XGBRegressor},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2604.04187}},
  note         = {Machine review of arXiv:2604.04187}
}

abstract

Building on the work of E. L. Medeiros [1] and our previous study [2], we generalize the alpha-nucleus nonlocality effect to odd-A and odd-odd nuclei within the two-potential approach (TPA) framework. The coordinate-dependent parameters introduced by this nonlocality are optimized using an advanced gradient boosting regression model. This improved TPA is applied to calculate the $\alpha$-decay half-lives of 599 nuclei with $Z = 52-118$, yielding a root-mean-square (RMS) deviation that is 74.8$\%$ lower than that of the original TPA. Subsequently, we employ the improved TPA, along with the DZR [3] and MUDL [4]models, to predict alpha-decay half-lives for 142 superheavy nuclei with $Z = 117-120$. The predictions from all three models are in close agreement, with the results from our improved TPA and the DZR model being nearly identical.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

2 extracted references · 1 linked inside Pith

[1]

Oxford University Press, (2009) [Ara64] Araki, H.,Von Neumann algebras of local observables for free scalar fieldJ

[Ara09] Araki, H.,Mathematical Theory of Quantum Fields. Oxford University Press, (2009) [Ara64] Araki, H.,Von Neumann algebras of local observables for free scalar fieldJ. Math. Phys., 5 (1) pp. 1-13 (1964) [BuSi98] Busch, P. and Singh, J. ,L¨ uders theorem for unsharp quantum measurements. Physics Letters A 249, (1998) 10-12. [Bus99] Busch, P.,Unsharp l...

arXiv doi:10.1007/s11005-023-01746-z 2009
[2]

(eds)Advances in Algebraic Quantum Field Theory.Mathematical Physics Studies

[KhMo15] Khavkine, I., Moretti, V.,Algebraic QFT in Curved Spacetime and Quasifree Hadamard States: An Introduction.In: Brunetti, R., Dappiaggi, C., Fredenhagen, K., Yngvason, J. (eds)Advances in Algebraic Quantum Field Theory.Mathematical Physics Studies. Springer, Cham. (2015) [LaSc65] Langerholc, J. and Schroer, B.On the Structure of the von Neumann Al...

doi:10.1007/s11005-023-01689-5 2015

Pith tools

Reviewed July 13, 2026 · model on record in the stance chip above.

[1] [1]

Oxford University Press, (2009) [Ara64] Araki, H.,Von Neumann algebras of local observables for free scalar fieldJ

[Ara09] Araki, H.,Mathematical Theory of Quantum Fields. Oxford University Press, (2009) [Ara64] Araki, H.,Von Neumann algebras of local observables for free scalar fieldJ. Math. Phys., 5 (1) pp. 1-13 (1964) [BuSi98] Busch, P. and Singh, J. ,L¨ uders theorem for unsharp quantum measurements. Physics Letters A 249, (1998) 10-12. [Bus99] Busch, P.,Unsharp l...

arXiv doi:10.1007/s11005-023-01746-z 2009

[2] [2]

(eds)Advances in Algebraic Quantum Field Theory.Mathematical Physics Studies

[KhMo15] Khavkine, I., Moretti, V.,Algebraic QFT in Curved Spacetime and Quasifree Hadamard States: An Introduction.In: Brunetti, R., Dappiaggi, C., Fredenhagen, K., Yngvason, J. (eds)Advances in Algebraic Quantum Field Theory.Mathematical Physics Studies. Springer, Cham. (2015) [LaSc65] Langerholc, J. and Schroer, B.On the Structure of the von Neumann Al...

doi:10.1007/s11005-023-01689-5 2015