Pith. sign in

REVIEW 3 major objections 3 minor 56 references

Bench2ADVLM: A Closed-Loop Benchmark for Vision-language Models in Autonomous Driving

T0 review · 3 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Bench2ADVLM, a closed-loop benchmark that drives vision-language models in simulation and on real vehicles, reports that current ADVLMs perform worse interactively than open-loop tests suggest.

desk verdict A plausible and potentially valuable closed-loop evaluation framework for ADVLMs, but the abstract alone does not support the headline empirical claim and the interpreter-VLM confound is unresolved. read the letter →

arxiv 2508.02028 v2 pith:UYMMIDOL submitted 2025-08-04 cs.CV

Tianyuan Zhang , Ting Jin , Lu Wang , Jiangfan Liu , Siyuan Liang , Mingchuan Zhang , Aishan Liu , Xianglong Liu This is my paper · ORCID
classification cs.CV
keywords vision-languagemodelsautonomousdrivingclosed-loopevaluationbenchmarkscenariogenerationsafety-criticalscenariossimulation-to-realitytransferself-reflection
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 argues that the usual way of testing vision-language models for autonomous driving — feeding static images or short clips and scoring the output — misses what matters most: how a model behaves when it must react to the consequences of its own decisions. To fix this, it introduces Bench2ADVLM, a hierarchical closed-loop evaluation framework that assesses ADVLMs in real time across both simulation and physical vehicles. The framework routes each target model's high-level commands through a general-purpose interpreter VLM into standardized control actions, adds a physical control layer to actuate real cars, and uses a self-reflective scenario generator to hunt for failure modes. Across multiple state-of-the-art ADVLMs, the paper reports limited performance under closed-loop conditions, implying that open-loop scores overstate how ready these systems are to drive.

What carries the argument

The load-bearing mechanism is the dual-system adaptation architecture, inspired by dual-process theories of cognition: it lets any target ADVLM be tested in a common simulation environment by routing its high-level commands through a general-purpose interpreter VLM that standardizes them into mid-level control actions. Two companion mechanisms complete the pipeline: the physical control abstraction layer, which maps those mid-level actions into low-level actuation for real vehicles, and the self-reflective scenario generation module, which converts model behavior into new safety-critical scenarios. Together they establish a three-level hierarchy running from abstract reasoning, through mid-level simulation actions, to low-level real-world execution.

What would settle it

Run the same cohort of ADVLMs through Bench2ADVLM twice, swapping the interpreter VLM for a different general-purpose model, and compare scores and rankings; if rankings or overall scores shift substantially, the interpreter is a major contributor to measured performance and the claim that the framework exposes the target models' own closed-loop limitations is not yet established. A corroborating check is to score the interpreter's translations against ground-truth mid-level actions on a fixed set of scenarios.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is that closed-loop evaluation of ADVLMs is both necessary and now feasible, and that once it is applied, existing ADVLMs show limited performance. The framework connects three levels: the target ADVLM acts as the fast system producing high-level driving commands; a general-purpose interpreter VLM, the slow system, converts those heterogeneous commands into standardized mid-level control actions executable in simulation; and a physical control abstraction layer translates the same mid-level actions into low-level actuation signals, giving closed-loop testing on physical vehicles for the first time. A self-reflective scenario generation module probes the model's behavior, uncovers potential failure modes, and generates safety-critical scenarios automatically. The experiments across multiple state-of-the-art ADVLMs and platforms are presented as validating the framework's diagnostic strength rather than the models' competence.

Load-bearing premise

The headline conclusion rests on the assumption that the general-purpose interpreter VLM faithfully converts each target model's high-level commands into control actions without contributing its own perception or reasoning errors; if the interpreter errs, the benchmark measures a composite system rather than the target ADVLM alone.

Editorial extensions

If this is right

  • Open-loop static-input scores for ADVLMs cannot be treated as evidence of driving competence, because interactive feedback changes model behavior and reveals weaknesses that static tests miss.
  • The dual-system adaptation architecture lets new ADVLMs enter closed-loop testing without per-model simulator integration, so the pipeline can track progress across the model family over time.
  • Self-reflective scenario generation turns evaluation into a failure-discovery tool: safety-critical scenarios are produced automatically rather than hand-authored.
  • Physical-vehicle testing, claimed as a first for ADVLMs, makes it possible to check whether simulation results transfer to real actuation and real latency.
  • The reported underperformance of state-of-the-art ADVLMs in closed-loop settings argues for training and validating these models in interactive environments, not only on static datasets.

Reading between the lines

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

  • A testable extension the paper leaves open: replacing the interpreter VLM with a different but similarly capable model and checking whether model rankings move would show how much of the measured deficiency belongs to the target model rather than to the interpreter.
  • Because the scenario generator loops model behavior back into scenario design, the same machinery could be turned into an adversarial training loop, using discovered failure scenarios to fine-tune the target model; the paper presents the loop as diagnostic, not generative.
  • The fast/slow split suggests a division-of-labor hypothesis the paper does not test: a strong interpreter may mask weaknesses in the target model's own perception, so reporting interpreter-independent diagnostics would sharpen the benchmark's conclusions.
  • A natural neighbouring application is to use the same three-level pipeline to evaluate closed-loop recovery behavior, such as responses after a mis-perceived sign or a near-miss, since interactive resilience rather than average accuracy is what the framework claims to isolate.
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 / 3 minor

Summary. The submission consists of an abstract for 'Bench2ADVLM', a proposed closed-loop, hierarchical evaluation framework for vision-language models in autonomous driving (ADVLMs), followed by a full-text document that is an unrelated paper on a third-order weighted essentially non-oscillatory compact least-squares (WCLS) scheme for hyperbolic conservation laws. The abstract describes a dual-system adaptation architecture in which target ADVLMs emit high-level commands that a general-purpose VLM translates into standardized mid-level actions; a physical control abstraction layer for real-vehicle execution; and a self-reflective scenario-generation module. It claims that experiments across diverse scenarios, state-of-the-art ADVLMs, and physical platforms validate the framework and reveal limited closed-loop performance of existing ADVLMs. As submitted, no methodology, experimental protocol, metrics, baselines, results, or implementation details for Bench2ADVLM are present beyond the abstract.

Significance. If properly developed and validated, the proposed framework would address a recognized gap in ADVLM evaluation, which is currently dominated by open-loop, static benchmarks. The combination of simulation-based closed-loop testing, a physical-vehicle control abstraction, and automated failure-mode scenario generation is potentially valuable to both the autonomous-driving and multimodal-model communities. However, the submission provides no verifiable evidence for any of these components: there is no code, no dataset, no machine-checked derivation, and no experimental data. The significance is therefore entirely conditional on a manuscript that is not actually included.

major comments (3)
  1. [Full Text (entire submission)] The full-text body is arXiv:2508.02033v1, a numerical-analysis paper on WCLS schemes for hyperbolic conservation laws, and it contains no description of Bench2ADVLM, its modules, or its experiments. Consequently, the abstract's central claim that 'Experiments ... validate the diagnostic strength of our framework' is unsupported by any section, equation, table, or figure in the submission. This is not a local presentation issue; the actual content needed for evaluation is absent, so the manuscript cannot be assessed on its merits.
  2. [Abstract, dual-system adaptation] The benchmark's diagnostic claim depends on the dual-system adaptation architecture, in which target-model high-level commands are converted by a general-purpose interpreter VLM into standardized mid-level actions. For this to measure the target ADVLM, the interpreter must be a behavior-preserving channel, but the abstract reports no oracle or replay baseline, no interpreter ablation, and no calibration against direct command execution. Without such a study, closed-loop scores conflate target-ADVLM capability with interpreter compatibility, so the headline claim that 'existing ADVLMs still exhibit limited performance under closed-loop conditions' is not established.
  3. [Abstract, experimental claims] The abstract's experimental claim lacks all protocol-level detail: metric definitions (e.g., task completion, safety violations, intervention rate), scenario taxonomy, number of runs, error bars, and comparison baselines are not given. Even if the body were the correct paper, a results claim of this breadth would need at least such information to support the stated conclusion.
minor comments (3)
  1. [Abstract] The phrase 'enabling, for the first time, closed-loop testing of ADVLMs on physical vehicles' should be qualified with a comparison to prior real-vehicle or closed-loop VLM evaluation work, otherwise the novelty claim is hard to verify.
  2. [Abstract] The terms 'fast system' and 'slow system' are used without formal definitions; the interface between the target model and the interpreter VLM (input/output vocabulary, prompting, failure handling) should be specified.
  3. [General] No URL, repository, or supplementary material is provided for the benchmark, which would be needed for reproducibility once the correct manuscript text is available.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the WCLS3 scheme's derivation is self-contained; the only self-citations are non-load-bearing.

full rationale

The derivation chain of the WCLS3 scheme is self-contained. The compact least-squares reconstruction is reviewed in Eqs. (3)-(10); the nonlinear weighting and dissipation terms are defined in Eqs. (11)-(17); the linear weights are optimized within the paper via the spectral optimization problem (Eqs. (19)-(22)) and tabulated in Table 1, even though the same optimization was originally reported in the authors' prior work [39]; the shock detector is defined in Eqs. (31)-(36) and its asymptotic behavior is proved in Theorem 1 using lemmas from the external reference [43]. Numerical experiments compare against independent baselines (WENO3-JS, WENO5-Z, CWENO3, CWENO5) on standard benchmarks and are not used to fit a parameter that is then reported as a prediction. The only self-citations, e.g., [39] for the optimized linear weights and [41] for an SSP-RK variant, are not load-bearing because the relevant equations and values appear in the present text. Note that the supplied full text corresponds to the CFD paper on the WCLS scheme rather than the Bench2ADVLM abstract, and the analysis above applies to the actual derivation text provided. No circular step that reduces a prediction to its input by construction was found.

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

The abstract exposes no fitted parameters and no invented physical entities. The framework's validity rests on several domain assumptions that are asserted rather than demonstrated in the abstract: the interpreter's fidelity, the superiority of closed-loop evaluation, and the representativeness of self-generated scenarios.

assumptions (3)
  • domain assumption A general-purpose VLM can faithfully translate high-level driving commands from arbitrary target ADVLMs into standardized mid-level control actions without distorting the target model's intent.
    This is the core of the dual-system adaptation architecture described in the abstract. If the interpreter is imperfect, the benchmark measures a composite of the target model and the interpreter, not the target model alone.
  • domain assumption Closed-loop simulation and physical-vehicle testing provide a valid and more informative measure of ADVLM performance than the open-loop static evaluation used in prior protocols.
    The entire motivating premise of the benchmark rests on this comparability. The abstract asserts open-loop protocols 'neglect interactive behavior' and treats closed-loop as 'more realistic and informative,' but provides no evidence in the abstract for this validity claim.
  • domain assumption The self-reflective scenario generation module produces safety-critical scenarios that are representative of real-world failure modes without biasing the evaluation.
    The abstract says the module 'automatically explores model behavior and uncovers potential failure modes,' but does not describe how scenario difficulty is controlled or how selection bias is avoided, which would affect the reported performance of ADVLMs.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Bench2ADVLM: A Closed-Loop Benchmark for Vision-language Models in Autonomous Driving." pith.science (2026). https://pith.science/paper/UYMMIDOL

@misc{pith2026250802028,
  author       = {Pith},
  title        = {Pith review of: Bench2ADVLM: A Closed-Loop Benchmark for Vision-language Models in Autonomous Driving},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UYMMIDOL}},
  note         = {Machine review of arXiv:2508.02028}
}
read the original abstract

Vision-Language Models (VLMs) have recently emerged as a promising paradigm in autonomous driving (AD). However, current performance evaluation protocols for VLM-based AD systems (ADVLMs) are predominantly confined to open-loop settings with static inputs, neglecting the more realistic and informative closed-loop setting that captures interactive behavior, feedback resilience, and real-world safety. To address this, we introduce Bench2ADVLM, a unified hierarchical closed-loop evaluation framework for real-time, interactive assessment of ADVLMs across both simulation and physical platforms. Inspired by dual-process theories of cognition, we first adapt diverse ADVLMs to simulation environments via a dual-system adaptation architecture. In this design, heterogeneous high-level driving commands generated by target ADVLMs (fast system) are interpreted by a general-purpose VLM (slow system) into standardized mid-level control actions suitable for execution in simulation. To bridge the gap between simulation and reality, we design a physical control abstraction layer that translates these mid-level actions into low-level actuation signals, enabling, for the first time, closed-loop testing of ADVLMs on physical vehicles. To enable more comprehensive evaluation, Bench2ADVLM introduces a self-reflective scenario generation module that automatically explores model behavior and uncovers potential failure modes for safety-critical scenario generation. Overall, Bench2ADVLM establishes a hierarchical evaluation pipeline that seamlessly integrates high-level abstract reasoning, mid-level simulation actions, and low-level real-world execution. Experiments on diverse scenarios across multiple state-of-the-art ADVLMs and physical platforms validate the diagnostic strength of our framework, revealing that existing ADVLMs still exhibit limited performance under closed-loop conditions.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

56 extracted references · 52 canonical work pages

  1. [1]

    X.-D. Liu, S. Osher, T. Chan, Weighted Essentially Non-oscillatory Schemes, Journal of Computational Physics 115 (1) (1994) 200–212

    work page 1994

  2. [2]

    Jiang, C.-W

    G.-S. Jiang, C.-W. Shu, Efficient implementation of weighted ENO schemes, Journal of Computational Physics 126 (1) (1996) 202–228

    work page 1996

  3. [3]

    C.-W. Shu, Essentially non-oscillatory and weighted essentially non-oscillatory schemes for hyperbolic conservation laws, in: Advanced Numerical Approximation of Nonlinear Hyperbolic Equations: Lectures given at the 2nd Session of the Centro Internazionale Matematico Estivo (CIME) held in Cetraro, Italy, June 23–28, 1997, Springer, 2006, pp. 325–432

    1997

  4. [4]

    D. Levy, G. Puppo, G. Russo, Compact central WENO schemes for multidimensional conservation laws, SIAM Journal on Scientific Com- puting 22 (2) (2000) 656–672

    work page 2000

  5. [5]

    Qiu, C.-W

    J. Qiu, C.-W. Shu, On the construction, comparison, and local characteristic decomposition for high-order central WENO schemes, Journal of Computational Physics 183 (1) (2002) 187–209

    work page 2002

  6. [6]

    Capdeville, A central WENO scheme for solving hyperbolic conservation laws on non-uniform meshes, Journal of Computational Physics 227 (5) (2008) 2977–3014

    G. Capdeville, A central WENO scheme for solving hyperbolic conservation laws on non-uniform meshes, Journal of Computational Physics 227 (5) (2008) 2977–3014

    work page 2008

  7. [7]

    H. Feng, F. Hu, R. Wang, A new mapped weighted essentially non-oscillatory scheme, Journal of Scientific Computing 51 (2) (2012) 449–473

    work page 2012

  8. [8]

    Baeza, R

    A. Baeza, R. Bürger, P. Mulet, D. Zorío, Central WENO schemes through a global average weight, Journal of Scientific Computing 78 (1) (2019) 499–530

    work page 2019

Show all 56 references
  1. [9]

    W.-G. Zeng, L. Liu, L.-J. Zeng, J.-H. Pan, J.-P. Yin, Y .-X. Ren, High-order alternative formulation of weighted essentially non-oscillatory scheme with minimized dispersion and controllable dissipation for compressible flows, International Journal for Numerical Methods in Flu...

    2025

  2. [10]

    G. B. Jacobs, W.-S. Don, A high-order WENO-Z finite difference based particle-source-in-cell method for computation of particle-laden flows with shocks, Journal of Computational Physics 228 (5) (2009) 1365–1379

    2009

  3. [11]

    Castro, B

    M. Castro, B. Costa, W. S. Don, High order weighted essentially non-oscillatory WENO-Z schemes for hyperbolic conservation laws, Journal of Computational Physics 230 (5) (2011) 1766–1792

    2011

  4. [12]

    Acker, R

    F. Acker, R. d. R. Borges, B. Costa, An improved WENO-Z scheme, Journal of Computational Physics 313 (2016) 726–753

    2016

  5. [13]

    A. K. Henrick, T. D. Aslam, J. M. Powers, Mapped weighted essentially non-oscillatory schemes: achieving optimal order near critical points, Journal of Computational Physics 207 (2) (2005) 542–567

    2005

  6. [14]

    Z. Hong, Z. Ye, X. Meng, A mapping-function-free WENO-M scheme with low computational cost, Journal of Computational Physics 405 (2020) 109145

    2020

  7. [15]

    R. Li, W. Zhong, An efficient mapped WENO scheme using approximate constant mapping, arXiv preprint arXiv:2102.00231 (2021)

    2021 arXiv

  8. [16]

    C. K. Tam, J. C. Webb, Dispersion-relation-preserving finite difference schemes for computational acoustics, Journal of Computational Physics 107 (2) (1993) 262–281

    1993

  9. [17]

    Cheong, S

    C. Cheong, S. Lee, Grid-optimized dispersion-relation-preserving schemes on general geometries for computational aeroacoustics, Journal of Computational Physics 174 (1) (2001) 248–276

    2001

  10. [18]

    Sun, Y .-X

    Z.-S. Sun, Y .-X. Ren, C. Larricq, S.-Y . Zhang, Y .-C. Yang, A class of finite difference schemes with low dispersion and controllable dissipation for DNS of compressible turbulence, Journal of Computational Physics 230 (12) (2011) 4616–4635

    2011

  11. [19]

    Z.-S. Sun, L. Luo, Y .-X. Ren, S.-Y . Zhang, A sixth order hybrid finite difference scheme based on the minimized dispersion and controllable dissipation technique, Journal of Computational Physics 270 (2014) 238–254

    2014

  12. [20]

    Li, Y .-X

    Y . Li, Y .-X. Ren, A scale-aware dispersion-relation-preserving finite difference scheme for computational aeroacoustics, Physics of Fluids 35 (3) (2023)

    2023

  13. [21]

    T. K. Sengupta, V . Lakshmanan, V . Vijay, A new combined stable and dispersion relation preserving compact scheme for non-periodic problems, Journal of Computational Physics 228 (8) (2009) 3048–3071

    2009

  14. [22]

    Popescu, W

    M. Popescu, W. Shyy, M. Garbey, Finite volume treatment of dispersion-relation-preserving and optimized prefactored compact schemes for wave propagation, Journal of Computational Physics 210 (2) (2005) 705–729

    2005

  15. [23]

    Q. Wang, Y . Ren, Z. Sun, Y . Sun, Low dispersion finite volume scheme based on reconstruction with minimized dispersion and controllable dissipation, Science China Physics, Mechanics and Astronomy 56 (2013) 423–431

    2013

  16. [24]

    S. K. Lele, Compact finite difference schemes with spectral-like resolution, Journal of Computational Physics 103 (1) (1992) 16–42

    1992

  17. [25]

    Cockburn, C.-W

    B. Cockburn, C.-W. Shu, Nonlinearly stable compact schemes for shock calculations, SIAM Journal on Numerical Analysis 31 (3) (1994) 607–627

    1994

  18. [26]

    N. A. Adams, K. Shariff, A high-resolution hybrid compact-ENO scheme for shock-turbulence interaction problems, Journal of Computa- tional Physics 127 (1) (1996) 27–51

    1996

  19. [27]

    Pirozzoli, Conservative hybrid compact-WENO schemes for shock-turbulence interaction, Journal of Computational Physics 178 (1) (2002) 81–117

    S. Pirozzoli, Conservative hybrid compact-WENO schemes for shock-turbulence interaction, Journal of Computational Physics 178 (1) (2002) 81–117

    2002

  20. [28]

    X. Deng, H. Maekawa, Compact high-order accurate nonlinear schemes, Journal of Computational Physics 130 (1) (1997) 77–91

    1997

  21. [29]

    Guo, Y .-f

    Y . Guo, Y .-f. Shi, Y .-m. Li, A fifth-order finite volume weighted compact scheme for solving one-dimensional Burgers’ equation, Applied Mathematics and Computation 281 (2016) 172–185

    2016

  22. [30]

    Y . Guo, Y . Shi, B. He, A weighted compact finite volume scheme for hyperbolic conservation laws, Communications on Applied Mathematics and Computation (2025) 1–25

    2025

  23. [31]

    Gamet, F

    L. Gamet, F. Ducros, F. Nicoud, T. Poinsot, Compact finite difference schemes on non-uniform meshes. Application to direct numerical simulations of compressible flows, International Journal for Numerical Methods in Fluids 29 (2) (1999) 159–191

    1999

  24. [32]

    Wang, Y .-X

    Q. Wang, Y .-X. Ren, An accurate and robust finite volume scheme based on the spline interpolation for solving the Euler and Navier-Stokes equations on non-uniform curvilinear grids, Journal of Computational Physics 284 (2015) 648–667

    2015

  25. [33]

    Huang, Y .-X

    W.-F. Huang, Y .-X. Ren, Q. Wang, X. Jiang, High resolution finite volume scheme based on the quintic spline reconstruction on non-uniform grids, Journal of Scientific Computing 74 (2018) 1816–1852

    2018

  26. [34]

    Wang, Y .-X

    Q. Wang, Y .-X. Ren, W. Li, Compact high order finite volume method on unstructured grids I: Basic formulations and one-dimensional schemes, Journal of Computational Physics 314 (2016) 863–882

    2016

  27. [35]

    Wang, Y .-X

    Q. Wang, Y .-X. Ren, W. Li, Compact high order finite volume method on unstructured grids II: Extension to two-dimensional Euler equations, Journal of Computational Physics 314 (2016) 883–908

    2016

  28. [36]

    Wang, Y .-X

    Q. Wang, Y .-X. Ren, J. Pan, W. Li, Compact high order finite volume method on unstructured grids III: Variational reconstruction, Journal of Computational physics 337 (2017) 1–26

    2017

  29. [37]

    J. Pan, Q. Wang, Y . Ren, et al., High-order compact finite volume methods on unstructured grids with adaptive mesh refinement for solving inviscid and viscous flows, Chinese Journal of Aeronautics 31 (9) (2018) 1829–1841

    2018

  30. [38]

    Huang, Y

    W. Huang, Y . Ren, G. Tu, Y . Xianxu, C. Jianqiang, An adaptive artificial viscosity method for quintic spline reconstruction scheme, Computers & Fluids 240 (2022) 105435

    2022

  31. [39]

    J. Pan, L. Li, J. Yin, W.-G. Zeng, Hybrid compact least-squares and central weighted essentially non-oscillatory schemes for hyperbolic conservation laws on structured curvilinear grids, Submission in Progress (2025)

    2025

  32. [40]

    Gottlieb, D

    S. Gottlieb, D. Ketcheson, C.-W. Shu, Strong stability preserving Runge-Kutta and multistep time discretizations, World Scientific, 2011

    2011

  33. [41]

    J. Pan, L. Li, Third-order unconditional positivity-preserving schemes for reactive flows keeping both mass and mole balance, Chinese Journal of Aeronautics 36 (11) (2023) 24–41

    2023

  34. [42]

    X. Hu, V . Tritschler, S. Pirozzoli, N. Adams, Dispersion-dissipation condition for finite difference schemes, arXiv preprint arXiv:1204.5088 (2012)

    2012 arXiv

  35. [43]

    Baeza, R

    A. Baeza, R. Bürger, P. Mulet, D. Zorío, An efficient third-order WENO scheme with unconditionally optimal accuracy, SIAM Journal on Scientific Computing 42 (2) (2020) A1028–A1051

    2020

  36. [44]

    Baeza, R

    A. Baeza, R. Bürger, P. Mulet, D. Zorío, On the efficient computation of smoothness indicators for a class of WENO reconstructions, Journal of Scientific Computing 80 (2019) 1240–1263. 38

    2019

  37. [45]

    Y . Li, C. Chen, Y . Ren, Z. Sun, Q. W ANG, The dispersion optimization and dissipation adjustment for high-order finite difference schemes, Acta Aerodynamica Sinica 39 (1) (2021) 138–156

    2021

  38. [46]

    Sanders, E

    R. Sanders, E. Morano, M.-C. Druguet, Multidimensional dissipation for upwind schemes: stability and applications to gas dynamics, Journal of Computational Physics 145 (2) (1998) 511–537

    1998

  39. [47]

    D. A. Bezgin, S. J. Schmidt, N. A. Adams, WENO3-NN: A maximum-order three-point data-driven weighted essentially non-oscillatory scheme, Journal of Computational Physics 452 (2022) 110920

    2022

  40. [48]

    G. A. Sod, A survey of several finite difference methods for systems of nonlinear hyperbolic conservation laws, Journal of computational physics 27 (1) (1978) 1–31

    1978

  41. [49]

    P. D. Lax, Weak solutions of nonlinear hyperbolic equations and their numerical computation, Communications on Pure and Applied Mathe- matics 7 (1) (1954) 159–193

    1954

  42. [50]

    Woodward, P

    P. Woodward, P. Colella, The numerical simulation of two-dimensional fluid flow with strong shocks, Journal of Computational Physics 54 (1) (1984) 115–173

    1984

  43. [51]

    C.-W. Shu, S. Osher, Efficient implementation of essentially non-oscillatory shock-capturing schemes, II, Journal of computational physics 83 (1) (1989) 32–78

    1989

  44. [52]

    V . A. Titarev, E. F. Toro, Finite-volume WENO schemes for three-dimensional conservation laws, Journal of Computational Physics 201 (1) (2004) 238–260

    2004

  45. [53]

    V . Daru, C. Tenaud, Evaluation of TVD high resolution schemes for unsteady viscous shocked flows, Computers & fluids 30 (1) (2000) 89–113

    2000

  46. [54]

    Weinan, C.-W

    E. Weinan, C.-W. Shu, A numerical resolution study of high order essentially non-oscillatory schemes applied to incompressible flow, Journal of Computational Physics 110 (1) (1994) 39–46

    1994

  47. [55]

    Wakimura, S

    H. Wakimura, S. Takagi, F. Xiao, Symmetry-preserving enforcement of low-dissipation method based on boundary variation diminishing principle, Computers & Fluids 233 (2022) 105227

    2022

  48. [56]

    V . Daru, C. Tenaud, Numerical simulation of the viscous shock tube problem by using a high resolution monotonicity-preserving scheme, Computers & Fluids 38 (3) (2009) 664–676. 39

    2009

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.