pith. sign in

arxiv: 2605.19170 · v1 · pith:C2HXVZ5Enew · submitted 2026-05-18 · 📊 stat.ML · cs.LG

Reducing Diffusion Model Memorization with Higher Order Langevin Dynamics

Benjamin Sterling , M\'onica F. Bugallo , Tom Tirer This is my paper

Pith reviewed 2026-05-20 07:07 UTC · model grok-4.3

classification 📊 stat.ML cs.LG
keywords diffusion modelsmemorizationhigher-order Langevin dynamicsscore functionregularizationgenerative modelsprivacy
0
0 comments X

The pith

Higher-order Langevin dynamics govern diffusion data trajectories with low-pass filtered scores, reducing memorization.

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

The paper seeks to establish that higher-order Langevin dynamics reduce the tendency of diffusion models to reproduce exact training samples. It does this by showing that the data variable evolves under a low-pass-filtered version of the learned score function, where the amount of smoothing grows with the model order. A sympathetic reader would care because exact memorization raises privacy and copyright risks when models generate new content. The analysis also covers the optimal empirical score and distribution collapse, with real-data experiments supporting the theory.

Core claim

Higher-order Langevin dynamics introduce auxiliary variables that can be viewed as velocity and acceleration when the data variable is treated as position. These variables impose extra dynamical constraints, so the data variable's updates are driven by a low-pass-filtered version of the score function whose smoothness increases with order. This regularization prevents the model from collapsing onto individual training points and thereby mitigates memorization.

What carries the argument

Higher-order Langevin dynamics whose auxiliary variables produce a low-pass-filtered score function whose cutoff sharpens with order.

If this is right

  • Higher model order produces smoother score-driven trajectories and therefore less exact memorization.
  • The optimal empirical score under HOLD does not collapse to training points.
  • Real-world experiments confirm that HOLD exhibits lower memorization than standard first-order diffusion.
  • The approach supplies a practical advantage for generating content without reproducing protected training instances.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same auxiliary-variable construction could be tested in other score-based samplers to control overfitting.
  • Frequency-domain analysis of the filtered score might reveal further connections to signal-processing ideas for generative models.
  • Extending the order to very high values could be checked on large-scale datasets to see whether the memorization benefit saturates.

Load-bearing premise

The auxiliary variables impose dynamical constraints that actually translate into low-pass filtering of the empirical score on the data variable.

What would settle it

An experiment in which raising the order of HOLD leaves the rate of exact training-sample reproduction unchanged would falsify the claimed link between filtering and reduced memorization.

Figures

Figures reproduced from arXiv: 2605.19170 by Benjamin Sterling, M\'onica F. Bugallo, Tom Tirer.

Figure 1
Figure 1. Figure 1: Magnitudes of the Fourier Transforms |H(iω)| for different HOLD diffusion model orders n, and the Ornstein–Uhlenbeck filter with ξ = 1. One may observe that the HOLD filters are better at attenuating higher frequencies than OU, while still allowing a wide band of lower frequencies. “smoothing” is manifested by h (n) t acting on the score function as a low-pass filter with stronger attenuation of high frequ… view at source ↗
Figure 2
Figure 2. Figure 2: Determinant of the inverse covariance matrix, assuming [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: CelebA FIDs and fraction memorized (Fmem) percentages by the number of training [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Nearest training neighbors for different models at [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: CIFAR-10 FIDs and fraction memorized (Fmem) percentages by the number of training [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Initial auxiliary variable comparisons for [PITH_FULL_IMAGE:figures/full_fig_p017_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Nearest training neighbors for different models at [PITH_FULL_IMAGE:figures/full_fig_p018_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Training losses for ntrain = 256 for the VPSDE and HOLD n = 2, 3 SDEs. C Additional CelebA images with FIDs and Fmem This appendix section presents more generated samples from the CelebA experiment for different number of training images ntrain. D Miscellaneous details For all experiments, a learning rate of 1 × 10−4 , the Adam optimizer, 32 base channels, a dropout rate of 0.1, and 1000 diffusion model st… view at source ↗
Figure 9
Figure 9. Figure 9: Celeba comparison for ntrain = 256 training samples at 1,000,000 training iterations. 19 [PITH_FULL_IMAGE:figures/full_fig_p019_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Celeba comparison for ntrain = 512 training samples at 1,000,000 training iterations. 20 [PITH_FULL_IMAGE:figures/full_fig_p020_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Celeba comparison for ntrain = 1024 training samples at 1,000,000 training iterations. 21 [PITH_FULL_IMAGE:figures/full_fig_p021_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Celeba comparison for ntrain = 2048 training samples at 1,000,000 training iterations. 22 [PITH_FULL_IMAGE:figures/full_fig_p022_12.png] view at source ↗
read the original abstract

Diffusion/score-based models have emerged as powerful generative models, capable of generating high-quality samples that mimic the training data distribution. However, it has been observed that they are prone to reproducing training samples-known as "memorization"-potentially violating copyright and privacy. In this paper, we study the effect of Higher-Order Langevin Dynamics (HOLD) on this phenomenon. HOLD diffusion processes introduce auxiliary variables; if the data variable is interpreted as "position," then the auxiliary variables can be interpreted as "velocity" and "acceleration," depending on the chosen order of the model. They were originally proposed based on the intuition that they regularize the trajectories of the data variable by implicitly imposing additional dynamical constraints. Our work provides, to our knowledge, the first theoretical characterization of the regularization effect of HOLD. Specifically, we show that in HOLD, the dynamics of the data variable are governed by a low-pass-filtered version of the learned score function, with smoothness increasing with the order of HOLD. We then analyze the optimal empirical score and the possibility of distribution collapse. Together, our results explain the mitigation of memorization as the model order increases. Finally, we present an empirical study on real-world data that supports our theory and highlights this distinct advantage of HOLD over standard diffusion in practice.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The manuscript claims that Higher-Order Langevin Dynamics (HOLD) reduces memorization in diffusion/score-based models. It provides the first theoretical characterization showing that the data variable dynamics are governed by a low-pass-filtered version of the learned score function, with smoothness increasing with HOLD order. It then analyzes the optimal empirical score and distribution collapse to explain the mitigation of memorization as model order increases, supported by an empirical study on real-world data.

Significance. If the explanatory chain holds, the work offers a principled, training-free approach to controlling memorization via sampling dynamics order, which is significant for privacy and copyright issues in generative models. The combination of a new theoretical regularization result with analysis of collapse modes and empirical validation is a strength.

major comments (1)
  1. [Abstract and theoretical characterization sections] The low-pass filtering result is derived for the learned score function, but the memorization mitigation and distribution-collapse argument rely on properties of the optimal empirical score. The manuscript does not explicitly bridge how the filtered learned score prevents the high-frequency collapse modes identified for the optimal case (see the paragraph on intuition for HOLD and the subsequent analysis of the optimal empirical score). This link is load-bearing for the central claim.
minor comments (2)
  1. [Empirical study section] The empirical study description would benefit from explicit reporting of quantitative metrics, baselines, and controls to allow direct verification of the claimed advantage over standard diffusion.
  2. [HOLD process definition] Notation for auxiliary variables (velocity/acceleration) and the precise definition of the low-pass filter could be introduced with a short diagram or expanded equation for clarity.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their constructive feedback and positive assessment of the significance of our work on Higher-Order Langevin Dynamics for reducing memorization. We address the major comment below and outline the revisions we will make to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract and theoretical characterization sections] The low-pass filtering result is derived for the learned score function, but the memorization mitigation and distribution-collapse argument rely on properties of the optimal empirical score. The manuscript does not explicitly bridge how the filtered learned score prevents the high-frequency collapse modes identified for the optimal case (see the paragraph on intuition for HOLD and the subsequent analysis of the optimal empirical score). This link is load-bearing for the central claim.

    Authors: We agree that the connection between the low-pass filtering result (which holds for the learned score) and the collapse analysis (which characterizes the optimal empirical score) should be made more explicit to support the central claim. The low-pass filtering theorem shows that HOLD dynamics are governed by a smoothed version of whatever score is provided, including the learned score. The optimal-score analysis identifies that high-frequency components in the score induce collapse and memorization. Because the learned score approximates the optimal empirical score in high-density regions (where generation occurs), the same smoothing attenuates those high-frequency modes for the learned score as well. We acknowledge that this approximation-based link is currently implicit. In the revised manuscript we will add a short bridging paragraph immediately after the optimal-score collapse analysis, explicitly stating that the low-pass operator applied to the learned score suppresses the identified collapse modes via the approximation property, thereby explaining the observed mitigation of memorization with increasing HOLD order. revision: yes

Circularity Check

0 steps flagged

No significant circularity in the derivation of HOLD low-pass filtering and memorization mitigation

full rationale

The paper derives the low-pass-filtered dynamics on the learned score directly from the structure of the higher-order auxiliary variables in HOLD, then performs a separate analysis of the optimal empirical score and distribution collapse to connect increasing order to reduced memorization. No equations or steps reduce by construction to fitted parameters, self-referential definitions, or load-bearing self-citations; the central claims rest on independent theoretical characterization of the regularization effect rather than renaming known results or smuggling ansatzes via prior work. The derivation chain is self-contained against external dynamical analysis and does not collapse to its inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Review is based solely on the abstract; no explicit free parameters, axioms, or invented entities are stated. The core intuition that auxiliary variables impose dynamical constraints is treated as a modeling choice rather than a derived result.

pith-pipeline@v0.9.0 · 5760 in / 1234 out tokens · 27140 ms · 2026-05-20T07:07:46.328226+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

63 extracted references · 63 canonical work pages · 1 internal anchor

  1. [1]

    2015 , eprint=

    Deep Unsupervised Learning using Nonequilibrium Thermodynamics , author=. 2015 , eprint=

    work page 2015

  2. [2]

    Advances in Neural Information Processing Systems , volume=

    Denoising Diffusion Probabilistic Models , author=. Advances in Neural Information Processing Systems , volume=

    work page

  3. [3]

    International Conference On Machine Learning , pages=

    Variational inference with normalizing flows , author=. International Conference On Machine Learning , pages=. 2015 , organization=

    work page 2015

  4. [4]

    Score-Based Generative Modeling with Critically-Damped

    Tim Dockhorn and Arash Vahdat and Karsten Kreis , booktitle=. Score-Based Generative Modeling with Critically-Damped

    work page

  5. [5]

    International Conference on Learning Representations , year=

    Score-Based Generative Modeling through Stochastic Differential Equations , author=. International Conference on Learning Representations , year=

    work page

  6. [6]

    Estimation of Non-Normalized Statistical Models by Score Matching , journal =

    Aapo Hyv. Estimation of Non-Normalized Statistical Models by Score Matching , journal =. 2005 , volume =

    work page 2005

  7. [7]

    Advances in Neural Information Processing Systems , volume=

    Score-based diffusion meets annealed importance sampling , author=. Advances in Neural Information Processing Systems , volume=

    work page

  8. [8]

    1998 , eprint=

    Annealed Importance Sampling , author=. 1998 , eprint=

    work page 1998

  9. [9]

    Subspace Diffusion Generative Models , ISBN=

    Jing, Bowen and Corso, Gabriele and Berlinghieri, Renato and Jaakkola, Tommi , year=. Subspace Diffusion Generative Models , ISBN=. doi:10.1007/978-3-031-20050-2_17 , booktitle=

    work page doi:10.1007/978-3-031-20050-2_17

  10. [10]

    2021 , eprint=

    Cascaded Diffusion Models for High Fidelity Image Generation , author=. 2021 , eprint=

    work page 2021

  11. [11]

    Advances in Neural Information Processing Systems , volume=

    Wavelet score-based generative modeling , author=. Advances in Neural Information Processing Systems , volume=

    work page

  12. [12]

    2021 , eprint=

    Maximum Likelihood Training of Score-Based Diffusion Models , author=. 2021 , eprint=

    work page 2021

  13. [13]

    Normalizing flow neural networks by

    Xu, Chen and Cheng, Xiuyuan and Xie, Yao , journal=. Normalizing flow neural networks by

    work page

  14. [14]

    Generative Modelling with Higher-Order

    Shi, Ziqiang and Liu, Rujie , journal=. Generative Modelling with Higher-Order

    work page

  15. [15]

    Stochastic Processes and their Applications , volume=

    Reverse-time diffusion equation models , author=. Stochastic Processes and their Applications , volume=. 1982 , publisher=

    work page 1982

  16. [16]

    Avoiding the

    Putzer, Eugene J , journal=. Avoiding the. 1966 , publisher=

    work page 1966

  17. [17]

    Särkkä and A

    S. Särkkä and A. Solin , title =

    work page

  18. [18]

    ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pages=

    Noisy Image Restoration Based on Conditional Acceleration Score Approximation , author=. ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pages=. 2024 , organization=

    work page 2024

  19. [19]

    Langwave: Realistic Voice Generation Based on High-Order

    Shi, Ziqiang and Liu, Rujie , booktitle=. Langwave: Realistic Voice Generation Based on High-Order. 2024 , organization=

    work page 2024

  20. [20]

    2022 , eprint=

    Classifier-Free Diffusion Guidance , author=. 2022 , eprint=

    work page 2022

  21. [21]

    International Conference On Machine Learning , pages=

    Deep unsupervised learning using nonequilibrium thermodynamics , author=. International Conference On Machine Learning , pages=. 2015 , organization=

    work page 2015

  22. [22]

    Heusel, Martin and Ramsauer, Hubert and Unterthiner, Thomas and Nessler, Bernhard and Hochreiter, Sepp , journal=

    work page

  23. [23]

    2009 , institution=

    Learning Multiple Layers of Features from Tiny Images , author=. 2009 , institution=

    work page 2009

  24. [24]

    Attention is All you Need , volume =

    Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and Kaiser, ukasz and Polosukhin, Illia , booktitle =. Attention is All you Need , volume =

    work page

  25. [25]

    First Conference on Language Modeling , year=

    Mamba: Linear-Time Sequence Modeling with Selective State Spaces , author=. First Conference on Language Modeling , year=

    work page

  26. [26]

    , booktitle=

    Sterling, Benjamin and Bugallo, Mónica F. , booktitle=. Critically-Damped Third-Order. 2025 , volume=

    work page 2025

  27. [27]

    The Thirteenth International Conference on Learning Representations , year=

    Underdamped Diffusion Bridges with Applications to Sampling , author=. The Thirteenth International Conference on Learning Representations , year=

    work page

  28. [28]

    Chatterji and Peter L

    Xiang Cheng and Niladri S. Chatterji and Peter L. Bartlett and Michael I. Jordan , booktitle=. Underdamped. 2017 , url=

    work page 2017

  29. [29]

    Advances in Neural Information Processing Systems , editor=

    Riemannian Score-Based Generative Modelling , author=. Advances in Neural Information Processing Systems , editor=. 2022 , url=

    work page 2022

  30. [30]

    The Thirteenth International Conference on Learning Representations , year=

    Trivialized Momentum Facilitates Diffusion Generative Modeling on Lie Groups , author=. The Thirteenth International Conference on Learning Representations , year=

    work page

  31. [31]

    Fran. Kinetic. AI for Accelerated Materials Design - ICLR 2025 , year=

    work page 2025

  32. [32]

    Progressive Growing of

    Tero Karras and Timo Aila and Samuli Laine and Jaakko Lehtinen , booktitle=. Progressive Growing of. 2018 , url=

    work page 2018

  33. [33]

    2024 , cdate=

    Raghav Singhal and Mark Goldstein and Rajesh Ranganath , title=. 2024 , cdate=

    work page 2024

  34. [34]

    The Eleventh International Conference on Learning Representations , year=

    Where to Diffuse, How to Diffuse, and How to Get Back: Automated Learning for Multivariate Diffusions , author=. The Eleventh International Conference on Learning Representations , year=

    work page

  35. [35]

    Goodfellow, Ian and Pouget-Abadie, Jean and Mirza, Mehdi and Xu, Bing and Warde-Farley, David and Ozair, Sherjil and Courville, Aaron and Bengio, Yoshua , title =. Commun. ACM , month = oct, pages =. 2020 , issue_date =. doi:10.1145/3422622 , abstract =

    work page doi:10.1145/3422622 2020

  36. [36]

    2nd International Conference on Learning Representations (ICLR) , year=

    Auto-Encoding Variational Bayes , author=. 2nd International Conference on Learning Representations (ICLR) , year=

    work page

  37. [37]

    Proceedings of the 33rd International Conference on Neural Information Processing Systems , articleno =

    Song, Yang and Ermon, Stefano , title =. Proceedings of the 33rd International Conference on Neural Information Processing Systems , articleno =. 2019 , publisher =

    work page 2019

  38. [38]

    Bugallo , year=

    Benjamin Sterling and Chad Gueli and Mónica F. Bugallo , year=. Critically-Damped Higher-Order. 2506.21741 , archivePrefix=

    work page arXiv

  39. [39]

    Why Diffusion Models Don

    Tony Bonnaire and Rapha. Why Diffusion Models Don. The Thirty-ninth Annual Conference on Neural Information Processing Systems , year=

    work page

  40. [40]

    The 29th International Conference on Artificial Intelligence and Statistics , year=

    Denoising Score Matching with Random Features: Insights on Diffusion Models From Precise Learning Curves , author=. The 29th International Conference on Artificial Intelligence and Statistics , year=

    work page

  41. [41]

    Bartlett and Philip M

    Peter L. Bartlett and Philip M. Long and Gábor Lugosi and Alexander Tsigler , title =. Proceedings of the National Academy of Sciences , volume =. 2020 , doi =. https://www.pnas.org/doi/pdf/10.1073/pnas.1907378117 , abstract =

    work page doi:10.1073/pnas.1907378117 2020

  42. [42]

    Random Features for Large-Scale Kernel Machines , url =

    Rahimi, Ali and Recht, Benjamin , booktitle =. Random Features for Large-Scale Kernel Machines , url =

    work page

  43. [43]

    Advances in Neural Information Processing Systems , editor=

    Elucidating the Design Space of Diffusion-Based Generative Models , author=. Advances in Neural Information Processing Systems , editor=. 2022 , url=

    work page 2022

  44. [44]

    ICML 2023 Workshop on Structured Probabilistic Inference & Generative Modeling , year=

    Diffusion probabilistic models generalize when they fail to memorize , author=. ICML 2023 Workshop on Structured Probabilistic Inference & Generative Modeling , year=

    work page 2023

  45. [45]

    The Twelfth International Conference on Learning Representations , year=

    Generalization in diffusion models arises from geometry-adaptive harmonic representations , author=. The Twelfth International Conference on Learning Representations , year=

    work page

  46. [46]

    Forty-first International Conference on Machine Learning , year=

    The emergence of reproducibility and consistency in diffusion models , author=. Forty-first International Conference on Machine Learning , year=

    work page

  47. [47]

    Transactions on Machine Learning Research , issn=

    On Memorization in Diffusion Models , author=. Transactions on Machine Learning Research , issn=

    work page

  48. [48]

    and Salimans, Tim and Poole, Ben and Ho, Jonathan , booktitle =

    Kingma, Diederik P. and Salimans, Tim and Poole, Ben and Ho, Jonathan , booktitle =. Variational Diffusion Models , url =

    work page

  49. [49]

    Defending Diffusion Models Against Membership Inference Attacks via Higher-Order Langevin Dynamics

    Benjamin Sterling and Yousef El-Laham and Mónica F. Bugallo , year=. Defending Diffusion Models Against Membership Inference Attacks via Higher-Order. 2509.14225 , archivePrefix=

    work page internal anchor Pith review Pith/arXiv arXiv

  50. [50]

    Proceedings of the 40th International Conference on Machine Learning , pages =

    Are Diffusion Models Vulnerable to Membership Inference Attacks? , author =. Proceedings of the 40th International Conference on Machine Learning , pages =. 2023 , editor =

    work page 2023

  51. [51]

    arXiv preprint arXiv:2305.14712 , year=

    On the generalization of diffusion model , author=. arXiv preprint arXiv:2305.14712 , year=

    work page arXiv

  52. [52]

    1997 , publisher=

    Signals & systems , author=. 1997 , publisher=

    work page 1997

  53. [53]

    Advances in neural information processing systems , volume=

    Generative modeling by estimating gradients of the data distribution , author=. Advances in neural information processing systems , volume=

    work page

  54. [54]

    32nd USENIX security symposium (USENIX Security 23) , pages=

    Extracting training data from diffusion models , author=. 32nd USENIX security symposium (USENIX Security 23) , pages=

    work page

  55. [55]

    Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    Diffusion art or digital forgery? investigating data replication in diffusion models , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    work page

  56. [56]

    Advances in neural information processing systems , volume=

    Gans trained by a two time-scale update rule converge to a local nash equilibrium , author=. Advances in neural information processing systems , volume=

    work page

  57. [57]

    2009 , publisher=

    Learning multiple layers of features from tiny images , author=. 2009 , publisher=

    work page 2009

  58. [58]

    Proceedings of the IEEE international conference on computer vision , pages=

    Deep learning face attributes in the wild , author=. Proceedings of the IEEE international conference on computer vision , pages=

    work page

  59. [59]

    Chenlin Meng and Yutong He and Yang Song and Jiaming Song and Jiajun Wu and Jun-Yan Zhu and Stefano Ermon , booktitle=

    work page

  60. [60]

    Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    Image restoration by denoising diffusion models with iteratively preconditioned guidance , author=. Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages=

    work page

  61. [61]

    Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

    Zero-shot image restoration using few-step guidance of consistency models (and beyond) , author=. Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

    work page

  62. [62]

    Advances in neural information processing systems , volume=

    Denoising diffusion restoration models , author=. Advances in neural information processing systems , volume=

    work page

  63. [63]

    arXiv preprint arXiv:2212.03221 , year=

    Adir: Adaptive diffusion for image reconstruction , author=. arXiv preprint arXiv:2212.03221 , year=

    work page arXiv