Pith. sign in

REVIEW 3 major objections 4 minor 69 references

PAL -- Parallel active learning for machine-learned potentials

T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read PAL is an MPI-based library that makes active learning for machine-learned potentials asynchronous by running labeling, training, generation, and prediction as concurrent kernels; the paper derives and demonstrates speedups of 2–3 when…

desk verdict A real software artifact whose abstract oversells it: PAL is a clean MPI-based active learning library, but the 'substantial speed-ups' claim rides on an idealized model rather than end-to-end measurements. read the letter →

arxiv 2412.00401 v1 pith:NPODFX3Q submitted 2024-11-30 cs.LG cond-mat.mtrl-scics.DCphysics.chem-phphysics.comp-ph

Chen Zhou , Marlen Neubert , Yuri Koide , Yumeng Zhang , Van-Quan Vuong , Tobias Schlöder , Stefanie Dehnen , Pascal Friederich This is my paper · ORCID
classification cs.LGcond-mat.mtrl-scics.DCphysics.chem-phphysics.comp-ph
keywords activelearningmachine-learnedpotentialsMPIparallelizationuncertaintyquantificationquery-by-committeeasynchronousworkflowsmoleculardynamicssurrogatemodels
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

The paper tries to show that the usual way of building datasets for machine-learned potentials is needlessly serial: label a few configurations with quantum chemistry, retrain, run more simulations, repeat. PAL is a Python library built on MPI that splits this loop into five kernels — prediction, generator, oracle, training, and a controller — and runs them asynchronously, so labeling and retraining no longer block exploration. The paper derives a runtime identity in which the parallel loop takes the maximum of the three per-module costs instead of their sum, giving speedups near 2 when oracle and training cost balance and near 3 when all three modules cost the same. It then demonstrates the workflow on excited-state photodynamics, hydrogen-atom transfer in proteins, bismuth clusters, and thermo-fluid surrogate modeling. The value of the claim is that active learning becomes a routine, hands-off, parallel workflow that uses HPC resources continuously rather than one task at a time.

What carries the argument

The controller kernel is the mechanism that carries the argument: split into Manager and Exchange sub-processes, it owns the only communication paths between generators, predictors, oracles, and trainers, so user code never sees MPI calls. Multiple instances of each kernel run with their own ranks; the Exchange sub-kernel keeps the high-frequency generator-prediction loop alive while the Manager runs the slower oracle-buffer and training-data-buffer loops. All data crosses MPI as one-dimensional numpy arrays so message sizes are fixed and predictable, and model weights are shipped from the training kernel to prediction replicas as flat arrays. The rate-limiting assumption is expressed in the equation $T_{\mathrm{parallel}} = \max(N/P \cdot t_{\mathrm{oracle}}, t_{\mathrm{train}}, t_{\mathrm{gen}})$, where $N$ is the number of samples to label, $P$ the number of parallel oracle workers, and $t_{\mathrm{oracle}}, t_{\mathrm{train}}, t_{\mathrm{gen}}$ the per-cycle costs of the oracle, the training step, and 1000 generator-predictor steps. That max identity is what turns the sum of three costs into a runtime dominated by the slowest partner.

What would settle it

Run an identical active learning task twice on the same hardware — once with PAL and once with a conventional serial loop that uses parallel oracles only — and measure wall-clock time to reach a fixed test accuracy. If, in a balanced-cost configuration where oracle, training, and generation each take roughly equal time, the PAL run is less than about 3x faster, or if MPI communication time is a measurable fraction of the module times during the run, the perfect-overlap assumption behind the speedup formula is violated.

Watch

Extended reading notes

Core claim

On its own terms, the central claim is that the active learning process for machine-learned potentials can be decomposed into four independent computational roles — prediction, configuration generation, ground-truth labeling, and model training — plus a central controller, and that this decomposition converts a sequential pipeline into a set of concurrently executing MPI processes. The controller gathers predictions from a committee of models, evaluates uncertainty, forwards the most informative geometries to the first available oracle, buffers labeled data, and only then pushes a batch to trainers; meanwhile generators keep proposing new geometries using the latest replicated weights. Quantitative support comes from the simplified identity $T_{\mathrm{parallel}} = \max(N/P \cdot t_{\mathrm{oracle}}, t_{\mathrm{train}}, t_{\mathrm{gen}})$ against $T_{\mathrm{serial}} = N/P \cdot t_{\mathrm{oracle}} + t_{\mathrm{train}} + t_{\mathrm{gen}}$, which yields speedup $1+P/N$ when oracle and training dominate equally, and speedup 3 when oracle, training, and generation each take the same time. The paper argues that because all resources stay busy in the parallel loop, the workflow also explores more diverse geometries and trains on more data than the serial version, not just the same work in less time.

Load-bearing premise

The speedup claims rest on the assumption that oracle labeling, model training, and generation-prediction overlap perfectly in time with negligible communication overhead; if real workloads serialize on data dependencies, blocking MPI calls, or competing for the same GPUs and memory, the derived speedups near 2–3 do not follow.

Editorial extensions

If this is right

  • In atomistic active learning, expensive ab initio oracles (TDDFT, DFT) no longer block molecular dynamics exploration: while one geometry is being labeled, generators continue proposing new geometries and trainers update weights, so more diverse configurations per wall-clock hour enter the dataset.
  • When oracle and training costs are comparable, the derived speedup approaches $1+P/N$ and reaches 2 when every oracle worker is busy; when oracle, training, and generation are balanced, the speedup approaches 3, independent of the absolute cost scale.
  • Because labeling and training are decoupled, users can combine cheap oracles (xTB) with expensive ones, use rolling training sets, or switch oracle levels without touching the communication layer.
  • The same five-kernel backbone transfers beyond atomistics, as shown with convolutional surrogates for thermo-fluid flow and particle swarm optimization as the generator.
  • The library currently assumes a single workload manager and fixed-size messages; when prediction inference drops below roughly 10 ms, communication becomes the limiting factor and the generator-prediction loop needs a tighter coupling.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Our inference: the same decoupling should apply to any active-learning loop with an expensive labeling step, including experimental oracles such as automated synthesis or photon measurements, because the controller makes no atomistic assumptions and the kernel interfaces are generic.
  • Our inference: a direct before/after benchmark with identical data and hardware would isolate how much of the observed speedup comes from the overlap identity versus from parallel oracles alone, a separation the paper does not make explicit.
  • Our inference: the MPI message-passing backbone could be replaced by shared-memory queues for the inference-rate-limited regime, an architectural variant suggested by the paper's stated 10 ms bottleneck.
  • Our inference: the controller's uncertainty-based buffer pruning could be combined with multi-fidelity oracles, where the controller sends cheap labels to the trainer for immediate retraining and expensive labels only for samples whose uncertainty survives a re-check after retraining.
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.

Referee Report

3 major / 4 minor

Summary. The manuscript introduces PAL, a Python/MPI library that modularizes active learning for machine-learned potentials into five concurrently running kernels (prediction, generator, training, oracle, controller). The authors claim that this asynchronous, decoupled architecture reduces computational overhead, improves scalability, and delivers substantial speed-ups on CPU and GPU hardware. The paper describes the kernel interfaces, presents a simple analytic speedup model in the Supporting Information, reports one communication-timing measurement, and qualitatively sketches four application examples (photodynamics, hydrogen-transfer reactions, inorganic clusters, and thermo-fluid optimization) whose quantitative results are said to be published elsewhere.

Significance. If the speedup claims were supported by empirical measurement, PAL would be a useful contribution: it addresses a real bottleneck in active learning for ML potentials, where sequential oracle labeling, training, and generation underutilize HPC resources. The modular design with MPI is a sound engineering approach, the code is open source, and the four application areas demonstrate genuine versatility beyond atomistic simulations. However, the paper's central quantitative claim—substantial speed-ups—rests on an idealized analytic model rather than on measured end-to-end comparisons, and the single reported timing does not establish a speedup. The framework's value is plausible, but the paper as submitted does not provide the evidence needed to support the headline claim.

major comments (3)
  1. [SI S2.1, Eqs. (1)-(4)] The speedup model assumes perfect overlap of oracle, training, and generation/prediction with zero communication cost and no resource contention. Under these assumptions, Eq. (4) is not a lower bound on speedup as claimed; for a fixed workload with real overheads, it is an idealized upper bound. The use-case estimates in SI S2.2 (factors of 2 and 3) therefore follow by construction from the definitions of T_serial and T_parallel and are not validated against any measured serial or parallel execution. The abstract's claim of 'substantial speed-ups' needs to be supported by an end-to-end wall-clock comparison between PAL and a serial active-learning workflow using the same kernels and workloads.
  2. [Section 3.1 and Section 4] The only quantitative timing reported in the paper is the 51.5 ms forward pass versus 4.27 ms MPI communication/trajectory propagation. This demonstrates low communication overhead for that particular configuration, but it is not a speedup measurement relative to a serial workflow, nor does it quantify the overlap of oracle and training with generation. Moreover, Section 4 concedes that when inference time is 10 ms or less, communication becomes a bottleneck, and that variable-size messages add overhead. These statements are in tension with the general speedup claim and need to be addressed by benchmark results that include fast-inference regimes.
  3. [Section 3, applications] The four application examples are described only qualitatively and delegate all quantitative accuracy and timing results to prior publications (refs. 45-48). The reader cannot verify from this manuscript that PAL accelerates these workflows, the accuracy of the resulting ML models, or the efficiency of the active-learning loop. At minimum, the paper should provide one complete quantitative case study (e.g., end-to-end PAL runtime, number of oracle calls, model accuracy) so that the speedup and effectiveness claims can be assessed.
minor comments (4)
  1. [Throughout] There are numerous typos, including 'origianl draft' in the author contributions, 'Communictation bottleneck' in Section 4, 'Rumtime' in SI S2, 'intinilized' in SI S6, and 'miminal' in Section 4. A careful proofread is needed.
  2. [Section 2 and SI S3] The text inconsistently refers to 'machine learning' processes as both 'ml_process' and 'training kernel'/'learning' in different places; please unify the terminology for kernel names and process counts.
  3. [SI S2.2, Use Case 2] The statement that 'the parallel and serial runtime are approximately the same, leading to no substantial speedup' (S ≈ 1) directly contradicts the general claim of substantial speedups; the authors should discuss this case explicitly in the main text so readers understand the conditions under which PAL is beneficial.
  4. [Section 4] The discussion of hardware support says PAL is only tested on Slurm with a single node type; this limitation is relevant for the scalability claims and should be mentioned in the abstract or conclusion as a boundary condition.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the SI S2 speedup model is an explicitly assumed arithmetic identity, and the cited application examples are external prior publications, so the central claims do not reduce by construction to their inputs.

full rationale

The paper's headline speedup claim is supported by the analytical model in SI S2.1, where Eq. (1) defines T_serial and Eq. (2) defines T_parallel under the explicitly stated assumptions of perfect parallelization and ignored communication overhead. Eq. (4) then computes speedup as their ratio. This is a straightforward algebraic consequence of the stated assumptions, not a fitted parameter renamed as a prediction; there is no subset of data whose fitting forces the reported speedup factors. The use-case values of 2 and 3 are arithmetic special cases of the model, not empirical measurements, so they cannot be circular in the sense of being equivalent to their own inputs. The four application scenarios are cited to prior publications (refs 45-48), some with overlapping authorship, but the present paper does not use those citations to justify a uniqueness claim or to forbid alternative interpretations; they are external, peer-reviewed works and are not the load-bearing derivation of PAL's parallel speedup. Section 4 explicitly concedes a communication bottleneck for very fast inference and variable-size messages, which qualifies the strength of the speedup claim but is a scoping and evidence-strength issue, not circularity. I therefore find no step in the derivation chain that reduces to its own inputs, so the circularity score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The central claim rests on an idealized parallel model rather than fitted constants: no free parameters are fit to data in this paper, and the user-defined uncertainty thresholds in the toy examples are not load-bearing. Three domain assumptions carry the argument: perfect module overlap with zero communication cost, decomposability of all modules into small balanced tasks, and the validity of the group's earlier application papers as evidence. No new physical entities are introduced.

assumptions (3)
  • domain assumption Ignoring communication overhead, the runtime of the parallel workflow equals the maximum of the module runtimes, T_parallel = max(N/P * t_oracle, t_train, t_gen) (SI S2.1, Eq. 2).
    The central speedup claims rest on this perfect-overlap model, which the paper does not validate with measured end-to-end timings.
  • domain assumption All modules can be decomposed into small pieces (single oracle call, one training epoch, short generator run) without losing efficiency, enabling load balancing (SI S2.2).
    The speedup of 2 to 3 depends on this decomposability; real workloads may have coarse-grained tasks or synchronization barriers.
  • domain assumption The four application successes are established by the prior works cited (refs. 45 to 48), which this paper does not reproduce.
    The paper uses these citations as evidence that PAL is effective in practice, but the underlying results and datasets are external.

how reviews work

0 comments
Cite this review

Pith. "Pith review of PAL -- Parallel active learning for machine-learned potentials." pith.science (2026). https://pith.science/paper/NPODFX3Q

@misc{pith2026241200401,
  author       = {Pith},
  title        = {Pith review of: PAL -- Parallel active learning for machine-learned potentials},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NPODFX3Q}},
  note         = {Machine review of arXiv:2412.00401}
}
read the original abstract

Constructing datasets representative of the target domain is essential for training effective machine learning models. Active learning (AL) is a promising method that iteratively extends training data to enhance model performance while minimizing data acquisition costs. However, current AL workflows often require human intervention and lack parallelism, leading to inefficiencies and underutilization of modern computational resources. In this work, we introduce PAL, an automated, modular, and parallel active learning library that integrates AL tasks and manages their execution and communication on shared- and distributed-memory systems using the Message Passing Interface (MPI). PAL provides users with the flexibility to design and customize all components of their active learning scenarios, including machine learning models with uncertainty estimation, oracles for ground truth labeling, and strategies for exploring the target space. We demonstrate that PAL significantly reduces computational overhead and improves scalability, achieving substantial speed-ups through asynchronous parallelization on CPU and GPU hardware. Applications of PAL to several real-world scenarios - including ground-state reactions in biomolecular systems, excited-state dynamics of molecules, simulations of inorganic clusters, and thermo-fluid dynamics - illustrate its effectiveness in accelerating the development of machine learning models. Our results show that PAL enables efficient utilization of high-performance computing resources in active learning workflows, fostering advancements in scientific research and engineering applications.

Figures

Figures reproduced from arXiv: 2412.00401 by the authors.

Figure 1
Figure 1. Comparison of (a) a conventional (serial) active learning and (b) our parallel active [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The computational architecture of the PAL workflow. Multiple boxes indicate paral [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Examples of PAL applications. a) Photodynamics simulations. b) Hydrogen atom [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Methods and data flow of PAL. Supporting information S1 Parallelization with MPI The PAL is designed in an object-oriented manner, wherein the implementation of each kernel is for a single corresponding process and the user-specified number of processes are generated f…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

69 extracted references · 66 canonical work pages

  1. [1]

    M. M. Bejani and M. Ghatee,Artificial Intelligence Review, 2021,54, 6391–6438

    work page 2021

  2. [2]

    M. I. Jordan and T. M. Mitchell,Science, 2015,349, 255–260

    work page 2015

  3. [3]

    Settles,Active learning literature survey, University of Wisconsin-Madison Department of Computer Sciences, 2009

    B. Settles,Active learning literature survey, University of Wisconsin-Madison Department of Computer Sciences, 2009

    work page 2009

  4. [4]

    Settles, Active learning and experimental design workshop in conjunction with AISTATS 2010, 2011, pp

    B. Settles, Active learning and experimental design workshop in conjunction with AISTATS 2010, 2011, pp. 1–18

    work page 2010

  5. [5]

    P. Ren, Y. Xiao, X. Chang, P.-Y. Huang, Z. Li, B. B. Gupta, X. Chen and X. Wang,ACM computing surveys (CSUR), 2021,54, 1–40

    work page 2021

  6. [6]

    Behler and M

    J. Behler and M. Parrinello,Physical review letters, 2007,98, 146401

    work page 2007

  7. [7]

    Bussi and A

    G. Bussi and A. Laio,Nature Reviews Physics, 2020,2, 200–212

    work page 2020

  8. [8]

    Friederich, F

    P. Friederich, F. Häse, J. Proppe and A. Aspuru-Guzik,Nature Materials, 2021,20, 750–761

    work page 2021

Show all 69 references
  1. [9]

    Behler,Chemical Reviews, 2021,121, 10037–10072

    J. Behler,Chemical Reviews, 2021,121, 10037–10072

    2021

  2. [10]

    V. L. Deringer, A. P. Bartók, N. Bernstein, D. M. Wilkins, M. Ceriotti and G. Csányi, Chemical Reviews, 2021,121, 10073–10141

    2021

  3. [11]

    Batzner, A

    S. Batzner, A. Musaelian, L. Sun, M. Geiger, J. P. Mailoa, M. Kornbluth, N. Molinari, T. E. Smidt and B. Kozinsky,Nature communications, 2022,13, 2453

    2022

  4. [12]

    Batatia, P

    I. Batatia, P. Benner, Y. Chiang, A. M. Elena, D. P. Kovács, J. Riebesell, X. R. Advincula, M. Asta, W. J. Baldwin, N. Bernsteinet al., arXiv preprint arXiv:2401.00096, 2023

    2023 arXiv

  5. [13]

    H. Chen, B. Roux and C. Chipot,Journal of Chemical Theory and Computation, 2023,19, 4414–4426

    2023

  6. [14]

    E. D. Donkor, A. Laio and A. Hassanali,Journal of Chemical Theory and Computation, 2023, 19, 4596–4605

    2023

  7. [15]

    Topel, A

    M. Topel, A. Ejaz, A. Squires and A. L. Ferguson,Journal of Chemical Theory and Compu- tation, 2023,19, 4654–4667

    2023

  8. [16]

    Sasmal, M

    S. Sasmal, M. McCullagh and G. M. Hocky,Journal of Chemical Theory and Computation, 2023, 19, 4427–4435

    2023

  9. [17]

    F. B. Akher, Y. Shu, Z. Varga and D. G. Truhlar,Journal of Chemical Theory and Compu- tation, 2023,19, 4389–4401

    2023

  10. [18]

    T. Hui, M. L. Descoteaux, J. Miao and Y.-S. Lin,Journal of Chemical Theory and Computa- tion, 2023,19, 4757–4769

    2023

  11. [19]

    Vandenhaute, M

    S. Vandenhaute, M. Cools-Ceuppens, S. DeKeyser, T. Verstraelen and V. Van Speybroeck, npj Computational Materials, 2023,9, 1–8

    2023

  12. [21]

    Vandermause, Y

    J. Vandermause, Y. Xie, J. S. Lim, C. J. Owen and B. Kozinsky,Nature Communications, 2022, 13, 5183

    2022

  13. [22]

    B. R. Duschatko, J. Vandermause, N. Molinari and B. Kozinsky,npj Computational Materials, 2024, 10, 1–12

    2024

  14. [23]

    J. S. Smith, B. Nebgen, N. Lubbers, O. Isayev and A. E. Roitberg,The Journal of chemical physics, 2018,148, 241733

    2018

  15. [24]

    Jinnouchi, K

    R. Jinnouchi, K. Miwa, F. Karsai, G. Kresse and R. Asahi,The Journal of Physical Chemistry Letters, 2020,11, 6946–6955. 12

    2020

  16. [25]

    Vandermause, S

    J. Vandermause, S. B. Torrisi, S. Batzner, Y. Xie, L. Sun, A. M. Kolpak and B. Kozinsky,npj Computational Materials, 2020,6, 1–11

    2020

  17. [26]

    W. Wang, T. Yang, W. H. Harris and R. Gómez-Bombarelli,Chemical Communications, 2020, 56, 8920–8923

    2020

  18. [27]

    J. Li, P. Reiser, B. R. Boswell, A. Eberhard, N. Z. Burns, P. Friederich and S. A. Lopez, Chemical Science, 2021,12, 5302–5314

    2021

  19. [28]

    Saleh, V

    Y. Saleh, V. Sanjay, A. Iske, A. Yachmenev and J. Küpper,The Journal of Chemical Physics, 2021, 155, 144109

    2021

  20. [29]

    S. J. Ang, W. Wang, D. Schwalbe-Koda, S. Axelrod and R. Gómez-Bombarelli,Chem, 2021, 7, 738–751

    2021

  21. [30]

    J. S. Smith, B. Nebgen, N. Mathew, J. Chen, N. Lubbers, L. Burakovsky, S. Tretiak, H. A. Nam, T. Germann, S. Fensin and K. Barros,Nature Communications, 2021,12, 1257

    2021

  22. [31]

    T. A. Young, T. Johnston-Wood, H. Zhang and F. Duarte,Physical Chemistry Chemical Physics, 2022,24, 20820–20827

    2022

  23. [32]

    D. E. Farache, J. C. Verduzco, Z. D. McClure, S. Desai and A. Strachan,Computational Materials Science, 2022,209, 111386

    2022

  24. [33]

    D. E. Kleiman and D. Shukla,Journal of Chemical Theory and Computation, 2023,19, 4377– 4388

    2023

  25. [34]

    Z. Li, H. Dong, J. Wang, L. Liu and J.-Y. Yang,International Journal of Heat and Mass Transfer, 2024,225, 125404

    2024

  26. [35]

    Zhang, V

    H. Zhang, V. Juraskova and F. Duarte,Nature Communications, 2024,15, 6114

    2024

  27. [36]

    Kahle and F

    L. Kahle and F. Zipoli,Physical Review E, 2022,105, 015311

    2022

  28. [37]

    Kulichenko, K

    M. Kulichenko, K. Barros, N. Lubbers, Y. W. Li, R. Messerly, S. Tretiak, J. S. Smith and B. Nebgen,Nature Computational Science, 2023,3, 230–239

    2023

  29. [38]

    D. W. Walker, Standards for message-passing in a distributed memory environment, 1992, https://www.osti.gov/biblio/7104668

    1992

  30. [39]

    Dalcin and Y.-L

    L. Dalcin and Y.-L. L. Fang,Computing in Science & Engineering, 2021,23, 47–54

    2021

  31. [40]

    Rogowski, S

    M. Rogowski, S. Aseeri, D. Keyes and L. Dalcin,IEEE Transactions on Parallel and Dis- tributed Systems, 2023,34, 611–622

    2023

  32. [41]

    K. T. Schütt, H. E. Sauceda, P.-J. Kindermans, A. Tkatchenko and K.-R. Müller,The Journal of Chemical Physics, 2018,148, 241722

    2018

  33. [42]

    Musaelian, S

    A. Musaelian, S. Batzner, A. Johansson, L. Sun, C. J. Owen, M. Kornbluth and B. Kozinsky, Nature Communications, 2023,14, 579

    2023

  34. [43]

    Batatia, D

    I. Batatia, D. P. Kovacs, G. N. C. Simm, C. Ortner and G. Csanyi, Advances in Neural Information Processing Systems, 2022

    2022

  35. [44]

    J. T. Ash and R. P. Adams,arXiv preprint arXiv:1910.08475, 2020

    1910 arXiv

  36. [45]

    C. Zhou, P. Kumar, D. Escudero and P. Friederich, AI for Accelerated Materials Design - NeurIPS 2023 Workshop, 2023

    2023

  37. [46]

    Riedmiller, P

    K. Riedmiller, P. Reiser, E. Bobkova, K. Maltsev, G. Gryn’ova, P. Friederich and F. Gräter, Chemical Science, 2024,15, 2518–2527

    2024

  38. [47]

    A. J. Kaithakkal, Y. Koide, M. Schniewind, P. Friederich and A. Stroh, International Heat Transfer Conference Digital Library, 2023

    2023

  39. [48]

    Koide, A

    Y. Koide, A. J. Kaithakkal, M. Schniewind, B. P. Ladewig, A. Stroh and P. Friederich,APL Machine Learning, 2024,2, 016108. 13

    2024

  40. [49]

    Pollice, P

    R. Pollice, P. Friederich, C. Lavigne, G. d. P. Gomes and A. Aspuru-Guzik,Matter, 2021,4, 1654–1682

    2021

  41. [50]

    Friederich, V

    P. Friederich, V. Gómez, C. Sprau, V. Meded, T. Strunk, M. Jenne, A. Magri, F. Symalla, A. Colsmann, M. Ruben and W. Wenzel,Advanced Materials, 2017,29, 1703505

    2017

  42. [51]

    Friederich, A

    P. Friederich, A. Fediai, S. Kaiser, M. Konrad, N. Jung and W. Wenzel,Advanced Materials, 2019, 31, 1808256

    2019

  43. [52]

    Reiser, M

    P. Reiser, M. Konrad, A. Fediai, S. Léon, W. Wenzel and P. Friederich,Journal of Chemical Theory and Computation, 2021,17, 3750–3759

    2021

  44. [53]

    Kunkel, J

    C. Kunkel, J. T. Margraf, K. Chen, H. Oberhofer and K. Reuter,Nature Communications, 2021, 12, 2422

    2021

  45. [54]

    Tan and D

    T. Tan and D. Wang,The Journal of Chemical Physics, 2023,158, 094102

    2023

  46. [55]

    H. M. Johnson, F. Gusev, J. T. Dull, Y. Seo, R. D. Priestley, O. Isayev and B. P. Rand, Journal of the American Chemical Society, 2024,146, 21583–21590

    2024

  47. [56]

    H. S. Seung, M. Opper and H. Sompolinsky, Proceedings of the Fifth Annual Workshop on Computational Learning Theory, New York, NY, USA, 1992, p. 287–294

    1992

  48. [57]

    C. Zapp, A. Obarska-Kosinska, B. Rennekamp, M. Kurth, D. M. Hudson, D. Mercadante, U. Barayeu, T. P. Dick, V. Denysenkov, T. Prisner, M. Bennati, C. Daday, R. Kappl and F. Gräter,Nature communications, 2020,11, 2315

    2020

  49. [58]

    Rennekamp, F

    B. Rennekamp, F. Kutzki, A. Obarska-Kosinska, C. Zapp and F. Gräter,Journal of Chemical Theory and Computation, 2020,16, 553–563

    2020

  50. [59]

    Ahlrichs, M

    R. Ahlrichs, M. Bär, M. Häser, H. Horn and C. Kölmel,Chemical Physics Letters, 1989,162, 165–169

    1989

  51. [60]

    Bannwarth, E

    C. Bannwarth, E. Caldeweyher, S. Ehlert, A. Hansen, P. Pracht, J. Seibert, S. Spicher and S. Grimme,WIREs Computational Molecular Science, 2021,11, e1493

    2021

  52. [61]

    Schopmans, P

    H. Schopmans, P. Reiser and P. Friederich,Digital Discovery, 2023,2, 1414–1424

    2023

  53. [62]

    Eberhardt,Surface Science, 2002,500, 242–270

    W. Eberhardt,Surface Science, 2002,500, 242–270

    2002

  54. [63]

    Chmiela, A

    S. Chmiela, A. Tkatchenko, H. E. Sauceda, I. Poltavsky, K. T. Schütt and K.-R. Müller, Science Advances, 2017,3, e1603015

    2017

  55. [64]

    Fukami, K

    K. Fukami, K. Fukagata and K. Taira,Journal of Fluid Mechanics, 2019,870, 106–120

    2019

  56. [65]

    Sabater, P

    C. Sabater, P. Stürmer and P. Bekemeyer,AIAA Journal, 2022,60, 5249–5261

    2022

  57. [66]

    Vinuesa and S

    R. Vinuesa and S. L. Brunton,arXiv preprint arXiv:2110.02085, 2021

    2021 arXiv

  58. [67]

    Duraisamy,Physical Review Fluids, 2021,6, 050504

    K. Duraisamy,Physical Review Fluids, 2021,6, 050504

    2021

  59. [68]

    Koide, A

    Y. Koide, A. J. Kaithakkal, J. Teufel, A. Stroh and P. Friederich,Under revision., 2024

    2024

  60. [69]

    Kennedy and R

    J. Kennedy and R. Eberhart, In Proceedings of the IEEE international conference on neural networks IV, 1995, pp. 1942–1948

    1995

  61. [70]

    r e s u l t _ d i r

    D. Müller,MA thesis, Institute of Fluid Mechanics, Karlsruhe Institute of Technology, Kaiser- straße 10, D-76131 Karlsruhe, 2021. 14 Figure 4: Methods and data flow of PAL. Supporting information S1 Parallelization with MPI The PAL is designed in an object-oriented manner, whe...

    2021

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.