Pith. sign in

REVIEW 3 major objections 8 minor 1 cited by

Model Tensor Planning

T0 review · 3 major / 8 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read Tensor sampling frees model-predictive control from local minima

desk verdict MTP is a genuinely practical extension of the authors' GTMP tensor-sampling idea into sampling-based MPC, with strong empirical results on contact-rich tasks; the catch is that the paper's theoretical guarantee, which is only for linear interpolation in an asymptotic limit, does not cover the spline variants that actually produce the wins. read the letter →

arxiv 2505.01059 v2 pith:2COGXYHN submitted 2025-05-02 cs.RO cs.AIcs.LGcs.SYeess.SY

An T. Le , Khai Nguyen , Minh Nhat Vu , João Carvalho , Jan Peters This is my paper · ORCID
classification cs.ROcs.AIcs.LGcs.SYeess.SY
keywords modelpredictivecontroltensorsamplingmultipartitegraphscross-entropymethodB-splineinterpolationAkimasplineexplorationexploitationtrade-offGPU-acceleratedplanning
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

Sampling-based model predictive control is powerful for nonlinear and contact-rich robots, but standard methods sample noise around a nominal trajectory, which traps them in local minima. This paper proposes Model Tensor Planning (MTP), which reformulates control sampling as tensor operations over randomized multipartite graphs: it uniformly samples control waypoints organized into layers, then interpolates smooth trajectories through them with B-splines or Akima splines. The paper claims this structured sampling achieves high-entropy, globally diverse control candidates at the same GPU cost as standard samplers, and that a beta-mixing blend of these global samples with local refinement in the Cross-Entropy update balances exploration and exploitation. If correct, it would make sampling MPC far more exploratory without dense reward shaping or differentiable dynamics, while remaining real-time and vectorized.

What carries the argument

The central machinery is tensor sampling over a random multipartite graph $G(M,N)$ control discretization: $M$ layers each containing $N$ control waypoints drawn uniformly from the control box, with complete directed connections between consecutive layers. Sampling with replacement from each layer forms a batch tensor $C \in \mathbb{R}^{B \times M \times n}$, which is interpolated along the time axis into full-horizon control trajectories using either linear segments, degree-$p$ B-splines (whose basis functions form a convex combination of waypoints), or Akima splines (piecewise cubics through the waypoints with local slope averaging). The graph structure converts path diversity into cheap tensor index operations, and the asymptotic path-coverage theorem converts the denseness of the graph in the limit into a promise that the sampler can reach any feasible control path.

What would settle it

Construct a control task with a narrow cost basin that can only be reached by a specific smooth trajectory whose waypoint sequence is not captured by the small $M,N$ grid. If MTP with $M=2-3$, $N=50-100$ and B-spline or Akima interpolation cannot concentrate samples there while a linear-interpolation variant or a denser graph can, the coverage claim for practical settings fails. More directly, compute the distance from a family of spline-interpolated paths to the nearest linear-interpolated graph path for the $M,N$ values used in Table 3; if the distance does not shrink toward zero or if the spline paths fail to intersect the low-cost region, the transfer assumption is false.

Watch

Extended reading notes

Core claim

The central claim is that a control trajectory can be treated as a path through a random multipartite graph whose layers are uniform random samples of control space, with each layer holding N candidate waypoints and edges connecting consecutive layers; sampling paths by picking one waypoint per layer and interpolating yields globally diverse, smooth control trajectories. The paper proves (Theorem 1) that for any bounded-variation control path, the distance from that path to the nearest graph path goes to zero as the number of layers M and waypoints per layer N both go to infinity, so the graph asymptotically covers the space of feasible control paths. Since each layer is sampled uniformly and independently, the distribution over graph paths has entropy $M \log N$, which grows without bound, meaning the sampler approaches maximum entropy over trajectory space. The method embeds these tensor-sampled trajectories into a modified Cross-Entropy update that softmax-weights elite candidates and mixes global tensor samples with local Gaussian samples in proportion $\beta$; experiments on dexterous manipulation and humanoid locomotion claim MTP matches or outperforms MPPI, Predictive Sampling, OpenAI-ES, and Diffusion Evolution in task success and control robustness, with ablations showing moderate $\beta$ values and B-spline/Akima interpolation are important.

Load-bearing premise

The asymptotic path-coverage guarantee is proven only for linear interpolation in the infinite limit, but the paper's reported success relies on B-spline and Akima interpolation on very small graphs ($M=2$ to $3$), so the argument assumes the linear denseness result transfers to spline-interpolated paths at practical sizes; that transfer is asserted rather than proved.

Editorial extensions

If this is right

  • MPC on GPU accelerators could maintain both global exploration and local refinement within a single batched rollout, replacing the usual choice between exploratory noise and exploitation.
  • Trajectory entropy becomes a tunable design quantity: increasing $M$ or $N$ raises the entropy of the candidate set, with the $M \log N$ closed form guiding how to spend a fixed rollout budget.
  • Because the sampler does not rely on gradients or dense reward shaping, it may extend to partially observed or contact-rich platforms where differentiable dynamics and informative costs are unavailable.
  • B-spline and Akima interpolation decouple the coarse waypoint exploration from the smoothness of the executed control, so exploration can be aggressive while the executed trajectory remains feasible.
  • The same tensorized graph structure, with vectorized JIT rollouts, suggests that online domain randomization over many dynamics models can be folded into the sampling loop without changing planning time.

Reading between the lines

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

  • The asymptotic path-coverage guarantee is proven only for linear interpolation and in the infinite limit; the paper's own experiments rely on B-spline and Akima interpolation at $M=2$ to $3$, so an implicit assumption is that finite-layer spline interpolation of graph paths still covers the cost-relevant control regions. A finite-$M$ or spline-specific denseness bound would directly test this.
  • The entropy identity $H = M \log N$ suggests a concrete scaling rule for sample efficiency: if the rollout budget $B$ is fixed, increasing $M$ multiplies the path space exponentially, so the effective coverage per sampled path shrinks; the paper's own recommendation to scale $N$ with $B$ is a partial answer, and a coverage-versus-budget analysis could make that tradeoff precise.
  • Because the tensor sampler is independent of the cost function and dynamics, it could be plugged into other trajectory optimizers or used as a high-entropy proposal distribution for diffusion-based or learned planners without changing the planner itself.
  • The $\beta$-mixing strategy could be made adaptive per timestep, using the entropy or cost variance of the batch to raise $\beta$ when the planner is stagnant and lower it when following an established nominal trajectory, which is a direct extension the paper leaves implicit.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 8 minor

Summary. The paper proposes Model Tensor Planning (MTP), a sampling-based MPC framework in which control candidates are generated by uniformly sampling waypoints on each layer of a random multipartite graph, interpolating the resulting point sequences (linearly, with B-splines, or with Akima splines), and mixing a fraction beta of these global samples with local Gaussian samples inside a CEM/MPPI-style update. The theoretical section claims asymptotic path coverage and maximum entropy of the tensor sampler as the number of graph layers M and waypoints per layer N tend to infinity. Experiments on six sim-to-sim MuJoCo tasks compare MTP variants against MPPI, Predictive Sampling, OpenAI-ES, and Diffusion Evolution, and additional ablations study beta, elite count, M, N, and B-spline degree, together with JAX/MuJoCo XLA wall-clock benchmarks.

Significance. If the theoretical claims were properly scoped, MTP would be a useful contribution to sampling-based MPC: the beta-mixing idea is simple and practical, the JAX implementation with JIT/vmap and batched rollouts is a genuine engineering strength, and the paper provides planning-time benchmarks and careful ablations of beta and E. The broad empirical comparison against both MPC and evolutionary baselines is also a strength. However, the theoretical support for the central 'high-entropy exploration' claim currently applies only to linear interpolation in an asymptotic regime, while the evaluated algorithm uses spline variants at small M; the maximum-entropy argument is also not formally valid. The empirical results may still stand on their own, but the paper's central theoretical narrative needs substantial revision.

major comments (3)
  1. [Appendix A.1, Theorem 1] The proof of Theorem 1 is internally inconsistent and incomplete. It states 'Since U is open' despite the paper's earlier assumption that the control space is compact; if u(t_i) lies on the boundary of U, an ambient ball is not contained in U (relative balls would fix this). More importantly, the theorem is stated as a deterministic limit, but the proof establishes only a probabilistic statement: the union bound gives success probability at least 1 - M e^{-cN}. For fixed M this tends to 1 as N grows, but under the double limit M,N -> infinity one must impose a rate such as N = omega(log M), which is absent. As written, the deterministic claim 'lim_{M,N->infty} min_g ||u-g||_infty = 0' is not established.
  2. [Section 3.1-3.2, Table 3] The path-coverage guarantee is proved only for linear interpolation, and Section 3.1 explicitly defers B-spline and Akima analyses to future work, yet Table 3 reports M=2-3 for the spline variants used in all main comparisons. For M=2, a degree-2 B-spline over two control points is an affine segment, and the Akima construction in Eq. (9) references m_{i-2}, so the M=2 Akima case is not even covered by Definition 3 as written. The support of tensor sampling is then a 2n-dimensional affine manifold inside the T x n trajectory space (e.g., dimension 58 versus 1160 for G1-Walk), not a dense subset. Consequently Theorem 1 does not support the exploration claim for the algorithm actually evaluated; the paper needs a finite-M or spline-specific denseness bound, or a clear restriction of the theoretical claims to the linear-interpolation asymptotic setting.
  3. [Appendix A.2, Eqs. (20)-(22)] The maximum-entropy argument compares incommensurable quantities. H(P_G) = M log N is the Shannon entropy of a finite categorical distribution over N^M index sequences, while H(P_MPPI) in Eq. (22) is a differential entropy in nats for a Gaussian over continuous control trajectories. The divergence of M log N as M,N grow reflects an increasingly fine discretization, not convergence to a maximum-entropy distribution over the continuous trajectory space, and there is no reference measure on the infinite-dimensional path space that would make the abstract's 'maximum entropy in the control trajectory space' claim meaningful. The maximum-entropy statement is therefore unsupported as written.
minor comments (8)
  1. [Algorithm 2, Line 7] For beta = 1, the expression B - P - 1 is negative, yet beta = 1 is used in the Navigation experiments and in the Fig. 5 sweep; the paper should specify max(B - P - 1, 0) or sample B - P - 1 local trajectories only when beta < 1.
  2. [Algorithm 2, Line 4] The control waypoint tensor is written as Q in R^{M x N x d}, but the control dimension is n throughout the paper; this should be R^{M x N x n}.
  3. [Section 2.1, Eq. (4)] The standard CEM covariance update divides by E - 1, which is undefined for E = 1; the ablation in Fig. 5 includes E = 1, so the preliminary equation should either state E > 1 or explain how the softmax/min-variance update in Algorithm 2 avoids this case.
  4. [Appendix A.2, Eq. (22)] If Sigma is the per-step covariance, the differential entropy of the Gaussian path distribution should contain (T/2) log det(Sigma), not T log det(Sigma); the factor of two affects the comparison and should be corrected.
  5. [Definition 1] Calling the graph a complete M-partite directed graph is imprecise: edges only connect consecutive layers, not arbitrary pairs across partitions; 'consecutive-layer complete directed bipartite graph' would be more accurate.
  6. [Section 3.1] The text says sampling without replacement is O(NM), but the number of combinatorial paths is N^M; the intended complexity is presumably O(N^M), and the notation should be fixed.
  7. [Section 3.2, Remark] The statement that for M,N -> infinity any g in F is also in G(M,N) is not well-defined, because the limiting object G(infinity, infinity) is not defined; the remark should be rephrased in terms of the convergence of the minimum distance.
  8. [Appendix A.5, Fig. 8] The y-axis label 'PustT' is a typo for 'PushT'.

Circularity Check

1 steps flagged · score 2.0 of 10

Maximum-entropy claim is true by construction; core path-coverage proof and MPC experiments are independent.

  1. self definitional [Appendix A.2, Eq. (21); Section 3.2 remark]
    "The uniform distribution over S is given by PG(τ) = 1/|S|, where |S| = N^M. Hence, the entropy of tensor sampling is H(PG) = −∑_{τ∈S}(1/N^M) log(1/N^M) = log(N^M) = M log N. Indeed, as M,N → ∞, the entropy H(PG) → ∞, and the distribution over sampled paths in G becomes maximum entropy over FT among all discrete path distributions."

    The 'maximum entropy' result is the uniform-per-layer construction restated: drawing each of the M layer indices uniformly from {1,...,N} makes the index sequence uniform over N^M sequences, and the entropy of that uniform law is M log N by definition. No optimization or independent characterization over control-trajectory space is performed; the entropy is fixed by the sampler's own design. The added denseness clause relies on Theorem 1, but the entropy identity itself is a self-definitional tautology rather than a derived guarantee. This does not infect the asymptotic path-coverage theorem or the empirical comparisons, which are separate.

full rationale

MTP's main derivation chain is not circular. The asymptotic path-coverage theorem (Theorem 1) is proved directly in Appendix A.1 by piecewise-linear approximation and a ball-coverage argument; it does not import its conclusion from Le et al. 2025 and is presented only as a guiding principle. The empirical claims are evaluated against MPPI, PS, OpenAI-ES, and DE on external MuJoCo tasks, with the same noise settings for MTP/MPPI/PS and with task-specific hyperparameters reported rather than hidden, so no fitted parameter is renamed as a prediction. The citations to Le et al. 2025 for random multipartite graphs and Akima-spline definitions are self-citations but non-load-bearing: the graph is redefined in Definition 1, the Akima spline is standard (Akima, 1974), and the paper's own theory covers the linear case. The only by-construction step is the 'maximum entropy' statement in Appendix A.2/Eq. (21), where uniform per-layer sampling makes H = M log N true by definition; calling this a derived maximum-entropy guarantee is a tautology, not an independent inference. This does not undermine the coverage theorem or the empirical wins. The theory-experiment gap (Theorem 1 is linear and asymptotic, while experiments use finite-M B-spline/Akima variants) is a correctness/scope risk, not circularity.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The central claim depends on several per-task hyperparameters, all reported in Table 3, and on domain assumptions about the control path class, compactness of the control space, availability of the dynamics model, and the effect of sampling with replacement. No new physical entities are introduced.

free parameters (7)
  • M (number of graph layers) = Task-dependent: 5 (Navigation), 2 (Crane, Cube-In-Hand, G1-Standup, G1-Walk, Walker), 3 (PushT)
    Chosen per task in Table 3; controls path complexity and directly enters the entropy formula H = M log N (Eq. 21).
  • N (waypoints per layer) = 30 (Navigation, Crane), 50 (Cube-In-Hand, PushT, Walker), 100 (G1-Standup, G1-Walk)
    Chosen per task in Table 3; larger N densifies the graph but increases memory and compute.
  • beta (mixing rate) = 1.0 (Navigation), 0.5 (Crane, Cube-In-Hand, PushT, Walker), 0.05 (G1-Standup), 0.02 (G1-Walk)
    Chosen per task in Table 3; balances global tensor samples versus local Gaussian samples.
  • E (number of elites) = 8 (Crane), 5 (Cube-In-Hand), 100 (G1-Standup, G1-Walk), 20 (PushT, Walker)
    Chosen per task in Table 3; affects the stability of the CEM-style update.
  • alpha (CEM smoothing factor) = 0.0 for most tasks, 0.1 (Cube-In-Hand), 0.5 (Walker)
    Chosen per task in Table 3; controls the exponential smoothing of mean and variance.
  • sigma_min (minimum variance) = 0.05 (Crane), 0.15 (Cube-In-Hand), 0.2 (G1-Standup), 0.1 (G1-Walk, PushT), 0.3 (Walker)
    Chosen per task in Table 3; clamps the covariance floor to avoid premature convergence.
  • lambda (softmax temperature) = not reported per task
    Algorithm 2 Line 11 uses lambda > 0 for softmax weighting, but no separate per-task values are given for MTP; it is an additional tuned constant.
assumptions (4)
  • domain assumption Control paths are uniformly continuous with bounded variation, defining the space F in Section 2.
    The theoretical results and the sampler are defined over this class; many robot controllers produce such paths, but the assumption excludes discontinuous control switches.
  • domain assumption Control space U is compact (control limits), making uniform sampling well-defined.
    Definition 1 samples u_j ~ Uniform(U); the proof of Theorem 1 uses both compactness and openness of U, which is inconsistent in Appendix A.1.
  • domain assumption A fixed dynamics model f(x,u) is available for rollouts.
    Algorithm 2 rolls out all candidates using f; the authors list this as a limitation in the Limitations section.
  • domain assumption Sampling with replacement from the graph does not meaningfully reduce path diversity.
    Section 3.1 argues duplicates are unlikely because N^M grows quickly; for small N and large B this assumption weakens.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Model Tensor Planning." pith.science (2026). https://pith.science/paper/2COGXYHN

@misc{pith2026250501059,
  author       = {Pith},
  title        = {Pith review of: Model Tensor Planning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2COGXYHN}},
  note         = {Machine review of arXiv:2505.01059}
}
abstract

Sampling-based model predictive control (MPC) offers strong performance in nonlinear and contact-rich robotic tasks, yet often suffers from poor exploration due to locally greedy sampling schemes. We propose \emph{Model Tensor Planning} (MTP), a novel sampling-based MPC framework that introduces high-entropy control trajectory generation through structured tensor sampling. By sampling over randomized multipartite graphs and interpolating control trajectories with B-splines and Akima splines, MTP ensures smooth and globally diverse control candidates. We further propose a simple $\beta$-mixing strategy that blends local exploitative and global exploratory samples within the modified Cross-Entropy Method (CEM) update, balancing control refinement and exploration. Theoretically, we show that MTP achieves asymptotic path coverage and maximum entropy in the control trajectory space in the limit of infinite tensor depth and width. Our implementation is fully vectorized using JAX and compatible with MuJoCo XLA, supporting \emph{Just-in-time} (JIT) compilation and batched rollouts for real-time control with online domain randomization. Through experiments on various challenging robotic tasks, ranging from dexterous in-hand manipulation to humanoid locomotion, we demonstrate that MTP outperforms standard MPC and evolutionary strategy baselines in task success and control robustness. Design and sensitivity ablations confirm the effectiveness of MTP tensor sampling structure, spline interpolation choices, and mixing strategy. Altogether, MTP offers a scalable framework for robust exploration in model-based planning and control.

Figures

Figures reproduced from arXiv: 2505.01059 by the authors.

Figure 1
Figure 1. Comparison of MTP interpolation methods versus CEM on [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Illustration of different tensor interpolations on an evenly spaced graph with [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Motivation comparison of MTP methods versus baselines with [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Performance comparison of MTP variants against standard MPC methods (MPPI, PS) and evolu [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Mixing rate β and number of elites E sweep on PushT, G1-Walk tasks with B = 128 to investigate the algorithmic update rule Algorithm 2 Line 9-12. The heatmap indicates accumulated cost over timesteps at termination, and the heat value range is fixed for each task/row. …
Figure 6
Figure 6. Figure 6: Mixing scalar β sweep on PushT, Cube-In-Hand, G1-Standup environments with B = 128 to investigate the sensitivity of MTP on explorative level. The dashed line represents the successful bar. In Cube-In-Hand, some of the cost curves increase due to the cube falling out o…
Figure 7
Figure 7. Figure 7: MTP-Bspline degree ablation. In G1-Walk, the Unitree G1 controlled Bspline degrees all roughly fall at 500 time steps. B-spline Degree Ablation. We investigate the sensitivity of MTP performance over B-spline interpolation degrees. The MTP hyperparameters are the same …
Figure 8
Figure 8. Figure 8: Sweep M, N on Navigation environment with B = 256 to investigate the interplay between number of batch sample B, number of layer M, and number of control-waypoints per layer N. Each data point is the success rate over 4 seeds. The environment setting is as in Appendix …
Figure 9
Figure 9. Figure 9: Softmax weighting ablation on PushT environment. Both control update trajectories converge to near-zero means with large variance at 100 timesteps, signifying task completion. 24 [PITH_FULL_IMAGE:figures/full_fig_p024_9.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. MOSAIC: Skill-Centric Manipulation Planning with Physics Simulation

    cs.RO 2025-04 conditional novelty 6.0 of 10

    MOSAIC is a multi-directional skill-centric planner that seeds feasible local trajectories with generator skills, links them with connector skills, and uses a statistical oracle and physics simulation to guide the search.

Reference graph

Works this paper leans on

49 extracted references · 25 canonical work pages · cited by 1 Pith paper

  1. [1]

    A method of bivariate interpolation and smooth surface fitting based on local procedures

    Hiroshi Akima. A method of bivariate interpolation and smooth surface fitting based on local procedures. Communications of the ACM, 17 0 (1): 0 18--20, 1974

    work page 1974

  2. [2]

    Real-time whole-body control of legged robots with model-predictive path integral control

    Juan Alvarez-Padilla, John Z Zhang, Sofia Kwok, John M Dolan, and Zachary Manchester. Real-time whole-body control of legged robots with model-predictive path integral control. arXiv preprint arXiv:2409.10469, 2024

    arXiv 2024

  3. [3]

    Learning dexterous in-hand manipulation

    OpenAI: Marcin Andrychowicz, Bowen Baker, Maciek Chociej, Rafal Jozefowicz, Bob McGrew, Jakub Pachocki, Arthur Petron, Matthias Plappert, Glenn Powell, Alex Ray, et al. Learning dexterous in-hand manipulation. The International Journal of Robotics Research, 39 0 (1): 0 3--20, 2020

    2020

  4. [4]

    Storm: An integrated framework for fast joint-space model-predictive control for reactive manipulation

    Mohak Bhardwaj, Balakumar Sundaralingam, Arsalan Mousavian, Nathan D Ratliff, Dieter Fox, Fabio Ramos, and Byron Boots. Storm: An integrated framework for fast joint-space model-predictive control for reactive manipulation. In Conference on Robot Learning, pp.\ 750--759. PMLR, 2022

    work page 2022

  5. [5]

    Massively parallelizing the rrt and the rrt

    Joshua Bialkowski, Sertac Karaman, and Emilio Frazzoli. Massively parallelizing the rrt and the rrt. In 2011 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp.\ 3513--3518. IEEE, 2011

    work page 2011

  6. [6]

    JAX : composable transformations of P ython+ N um P y programs, 2018

    James Bradbury et al. JAX : composable transformations of P ython+ N um P y programs, 2018. URL http://github.com/jax-ml/jax

    work page 2018

  7. [7]

    Carvalho, A

    J. Carvalho, A. Le, P. Kicki, D. Koert, and J. Peters. Motion planning diffusion: Learning and adapting robot motion planning with diffusion models, 2024. URL https://arxiv.org/abs/2412.19948

    arXiv 2024

  8. [8]

    Motion planning diffusion: Learning and planning of robot motions with diffusion models

    Joao Carvalho, An T Le, Mark Baierl, Dorothea Koert, and Jan Peters. Motion planning diffusion: Learning and planning of robot motions with diffusion models. In 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp.\ 1916--1923. IEEE, 2023

    2023

Show all 49 references
  1. [9]

    Diffusion policy: Visuomotor policy learning via action diffusion

    Cheng Chi, Zhenjia Xu, Siyuan Feng, Eric Cousineau, Yilun Du, Benjamin Burchfiel, Russ Tedrake, and Shuran Song. Diffusion policy: Visuomotor policy learning via action diffusion. The International Journal of Robotics Research, pp.\ 02783649241273668, 2023

    2023

  2. [10]

    A tutorial on the cross-entropy method

    Pieter-Tjerk De Boer, Dirk P Kroese, Shie Mannor, and Reuven Y Rubinstein. A tutorial on the cross-entropy method. Annals of operations research, 134: 0 19--67, 2005

    2005

  3. [11]

    Package for calculating with b-splines

    Carl de Boor. Package for calculating with b-splines. SIAM Journal on Numerical Analysis, 14: 0 57, 10 1973. doi:10.1137/0714026

    1973 doi

  4. [12]

    The cma evolution strategy: A tutorial

    Nikolaus Hansen. The cma evolution strategy: A tutorial. arXiv preprint arXiv:1604.00772, 2016

    2016 arXiv

  5. [13]

    Denoising diffusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. Advances in neural information processing systems, 33: 0 6840--6851, 2020

    2020

  6. [14]

    Predictive sampling: Real-time behaviour synthesis with mujoco

    Taylor Howell, Nimrod Gileadi, Saran Tunyasuvunakool, Kevin Zakka, Tom Erez, and Yuval Tassa. Predictive sampling: Real-time behaviour synthesis with mujoco. arXiv preprint arXiv:2212.00541, 2022

    2022 arXiv

  7. [15]

    prrtc: Gpu-parallel rrt-connect for fast, consistent, and low-cost motion planning

    Chih H Huang, Pranav Jadhav, Brian Plancher, and Zachary Kingston. prrtc: Gpu-parallel rrt-connect for fast, consistent, and low-cost motion planning. arXiv preprint arXiv:2503.06757, 2025

    2025

  8. [16]

    Diffusionseeder: Seeding motion optimization with diffusion for rapid motion planning

    Huang Huang, Balakumar Sundaralingam, Arsalan Mousavian, Adithyavairavan Murali, Ken Goldberg, and Dieter Fox. Diffusionseeder: Seeding motion optimization with diffusion for rapid motion planning. arXiv preprint arXiv:2410.16727, 2024

    arXiv

  9. [17]

    Vp-sto: Via-point-based stochastic trajectory optimization for reactive robot behavior

    Julius Jankowski, Lara Bruderm \"u ller, Nick Hawes, and Sylvain Calinon. Vp-sto: Via-point-based stochastic trajectory optimization for reactive robot behavior. In 2023 IEEE International Conference on Robotics and Automation (ICRA), pp.\ 10125--10131. IEEE, 2023

    2023

  10. [18]

    Rrt-connect: An efficient approach to single-query path planning

    James J Kuffner and Steven M LaValle. Rrt-connect: An efficient approach to single-query path planning. In Proceedings 2000 ICRA. Millennium conference. IEEE international conference on robotics and automation. Symposia proceedings (Cat. No. 00CH37065), volume 2, pp.\ 995--100...

    2000

  11. [19]

    Hydrax: Sampling-based model predictive control on gpu with jax and mujoco mjx, 2024

    Vince Kurtz. Hydrax: Sampling-based model predictive control on gpu with jax and mujoco mjx, 2024. https://github.com/vincekurtz/hydrax

    2024

  12. [20]

    evosax: Jax-based evolution strategies

    Robert Tjarko Lange. evosax: Jax-based evolution strategies. In Proceedings of the Companion Conference on Genetic and Evolutionary Computation, pp.\ 659--662, 2023

    2023

  13. [21]

    Gpu parallelization of policy iteration rrt

    R Connor Lawson, Linda Wills, and Panagiotis Tsiotras. Gpu parallelization of policy iteration rrt. In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp.\ 4369--4374. IEEE, 2020

    2020

  14. [22]

    Accelerating motion planning via optimal transport

    An T Le, Georgia Chalvatzaki, Armin Biess, and Jan R Peters. Accelerating motion planning via optimal transport. Advances in Neural Information Processing Systems, 36: 0 78453--78482, 2023

    2023

  15. [23]

    Global tensor motion planning

    An T Le, Kay Hansel, Jo \ a o Carvalho, Joe Watson, Julen Urain, Armin Biess, Georgia Chalvatzaki, and Jan Peters. Global tensor motion planning. arXiv preprint arXiv:2411.19393, 2024

    2024 arXiv

  16. [24]

    Drop: Dexterous reorientation via online planning

    Albert H Li, Preston Culbertson, Vince Kurtz, and Aaron D Ames. Drop: Dexterous reorientation via online planning. arXiv preprint arXiv:2409.14562, 2024

    2024 arXiv

  17. [25]

    Stochastic mpc with offline uncertainty sampling

    Matthias Lorenzen, Fabrizio Dabbene, Roberto Tempo, and Frank Allg \"o wer. Stochastic mpc with offline uncertainty sampling. Automatica, 81: 0 176--183, 2017

    2017

  18. [26]

    Isaac gym: High performance gpu-based physics simulation for robot learning

    Viktor Makoviychuk, Lukasz Wawrzyniak, Yunrong Guo, Michelle Lu, Kier Storey, Miles Macklin, David Hoeller, Nikita Rudin, Arthur Allshire, Ankur Handa, et al. Isaac gym: High performance gpu-based physics simulation for robot learning. arXiv preprint arXiv:2108.10470, 2021

    2021 arXiv

  19. [27]

    Model predictive control: Recent developments and future promise

    David Q Mayne. Model predictive control: Recent developments and future promise. Automatica, 50 0 (12): 0 2967--2986, 2014

    2014

  20. [28]

    Flowmp: Learning motion fields for robot planning with conditional flow matching

    Khang Nguyen, An T Le, Tien Pham, Manfred Huber, Jan Peters, and Minh Nhat Vu. Flowmp: Learning motion fields for robot planning with conditional flow matching. arXiv preprint arXiv:2503.06135, 2025

    2025 arXiv

  21. [29]

    Gpu-based parallel collision detection for fast motion planning

    Jia Pan and Dinesh Manocha. Gpu-based parallel collision detection for fast motion planning. The International Journal of Robotics Research, 31 0 (2): 0 187--200, 2012

    2012

  22. [30]

    Kino-pax: Highly parallel kinodynamic sampling-based planner

    Nicolas Perrault, Qi Heng Ho, and Morteza Lahijanian. Kino-pax: Highly parallel kinodynamic sampling-based planner. IEEE Robotics and Automation Letters, 2025

    2025

  23. [31]

    Sampling-based model predictive control leveraging parallelizable physics simulations

    Corrado Pezzato, Chadi Salmi, Elia Trevisan, Max Spahn, Javier Alonso-Mora, and Carlos Hern \'a ndez Corbato. Sampling-based model predictive control leveraging parallelizable physics simulations. IEEE Robotics and Automation Letters, 2025

    2025

  24. [32]

    Sample-efficient cross-entropy method for real-time planning

    Cristina Pinneri, Shambhuraj Sawant, Sebastian Blaes, Jan Achterhold, Joerg Stueckler, Michal Rolinek, and Georg Martius. Sample-efficient cross-entropy method for real-time planning. In Conference on Robot Learning, pp.\ 1049--1065. PMLR, 2021

    2021

  25. [33]

    Evolution strategies as a scalable alternative to reinforcement learning

    Tim Salimans, Jonathan Ho, Xi Chen, Szymon Sidor, and Ilya Sutskever. Evolution strategies as a scalable alternative to reinforcement learning. arXiv preprint arXiv:1703.03864, 2017

    2017 arXiv

  26. [34]

    Proximal policy optimization algorithms, 2017

    John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Radford, and Oleg Klimov. Proximal policy optimization algorithms, 2017. URL https://arxiv.org/abs/1707.06347

    2017 arXiv

  27. [35]

    Curobo: Parallelized collision-free robot motion generation

    Balakumar Sundaralingam, Siva Kumar Sastry Hari, Adam Fishman, Caelan Garrett, Karl Van Wyk, Valts Blukis, Alexander Millane, Helen Oleynikova, Ankur Handa, Fabio Ramos, et al. Curobo: Parallelized collision-free robot motion generation. In 2023 IEEE International Conference o...

    2023

  28. [36]

    Motions in microseconds via vectorized sampling-based planning

    Wil Thomason, Zachary Kingston, and Lydia E Kavraki. Motions in microseconds via vectorized sampling-based planning. In 2024 IEEE International Conference on Robotics and Automation (ICRA), pp.\ 8749--8756. IEEE, 2024

    2024

  29. [37]

    Mujoco: A physics engine for model-based control

    Emanuel Todorov, Tom Erez, and Yuval Tassa. Mujoco: A physics engine for model-based control. In 2012 IEEE/RSJ international conference on intelligent robots and systems, pp.\ 5026--5033. IEEE, 2012

    2012

  30. [38]

    Gymnasium: A standard interface for reinforcement learning environments

    Mark Towers, Ariel Kwiatkowski, Jordan Terry, John U Balis, Gianluca De Cola, Tristan Deleu, Manuel Goul \ a o, Andreas Kallinteris, Markus Krimmel, Arjun KG, et al. Gymnasium: A standard interface for reinforcement learning environments. arXiv preprint arXiv:2407.17032, 2024

    2024 arXiv

  31. [39]

    Learning implicit priors for motion optimization

    Julen Urain, An T Le, Alexander Lambert, Georgia Chalvatzaki, Byron Boots, and Jan Peters. Learning implicit priors for motion optimization. In 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp.\ 7672--7679. IEEE, 2022

    2022

  32. [40]

    Mppi-generic: A cuda library for stochastic optimization

    Bogdan Vlahov, Jason Gibson, Manan Gandhi, and Evangelos A Theodorou. Mppi-generic: A cuda library for stochastic optimization. arXiv preprint arXiv:2409.07563, 2024

    2024

  33. [41]

    Inferring smooth control: Monte carlo posterior policy iteration with gaussian processes

    Joe Watson and Jan Peters. Inferring smooth control: Monte carlo posterior policy iteration with gaussian processes. In Conference on Robot Learning, pp.\ 67--79. PMLR, 2023

    2023

  34. [42]

    Natural evolution strategies

    Daan Wierstra, Tom Schaul, Tobias Glasmachers, Yi Sun, Jan Peters, and J \"u rgen Schmidhuber. Natural evolution strategies. The Journal of Machine Learning Research, 15 0 (1): 0 949--980, 2014

    2014

  35. [43]

    Model predictive path integral control: From theory to parallel computation

    Grady Williams, Andrew Aldrich, and Evangelos A Theodorou. Model predictive path integral control: From theory to parallel computation. Journal of Guidance, Control, and Dynamics, 40 0 (2): 0 344--357, 2017

    2017

  36. [44]

    Full-order sampling-based mpc for torque-level locomotion control via diffusion-style annealing

    Haoru Xue, Chaoyi Pan, Zeji Yi, Guannan Qu, and Guanya Shi. Full-order sampling-based mpc for torque-level locomotion control via diffusion-style annealing. arXiv preprint arXiv:2409.15610, 2024

    2024 arXiv

  37. [45]

    Covo-mpc: Theoretical analysis of sampling-based mpc and optimal covariance design

    Zeji Yi, Chaoyi Pan, Guanqi He, Guannan Qu, and Guanya Shi. Covo-mpc: Theoretical analysis of sampling-based mpc and optimal covariance design. In 6th Annual Learning for Dynamics & Control Conference, pp.\ 1122--1135. PMLR, 2024

    2024

  38. [46]

    Diffusion models are evolutionary algorithms

    Yanbo Zhang, Benedikt Hartl, Hananel Hazan, and Michael Levin. Diffusion models are evolutionary algorithms. arXiv preprint arXiv:2410.02543, 2024

    2024 arXiv

  39. [47]

    A simple decentralized cross-entropy method

    Zichen Zhang, Jun Jin, Martin Jagersand, Jun Luo, and Dale Schuurmans. A simple decentralized cross-entropy method. Advances in Neural Information Processing Systems, 35: 0 36495--36506, 2022

    2022

  40. [48]

    Chomp: Covariant hamiltonian optimization for motion planning

    Matt Zucker, Nathan Ratliff, Anca D Dragan, Mihail Pivtoraiko, Matthew Klingensmith, Christopher M Dellin, J Andrew Bagnell, and Siddhartha S Srinivasa. Chomp: Covariant hamiltonian optimization for motion planning. The International journal of robotics research, 32 0 (9-10): ...

    2013

  41. [49]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.