Pith.

Robotics · arXiv:2610.24579 v1 · submitted 17 Oct 2026 · 54 references

LIBERO-Dynamic: Benchmarking Robotic Manipulation in Dynamic Environments

Shaojie Ren, Jingwen Sun, Zezhi Liu, Hongjin Chen, Xin Jin, Wei Zou, Zhibo Chen

The paper says

“Most simulation benchmarks for robotic manipulation keep target objects static and treat policy inference as instantaneous. These assumptions simplify evaluation but hide how a moving target evolves during computation and how stale an observation is when an action is executed.”

more of the abstract

We introduce LIBERO-Dynamic, a dynamic extension of all 40 LIBERO tasks. It drives objects through contact friction along three trajectory families and speed tiers, while advancing the physical world according to calibrated, model-specific inference latency. The same mechanism supports training by replaying verified static demonstrations and compensating robot actions online for target displacement, producing 9,000 quality-filtered dynamic trajectories. We evaluate six pretrained policies on 1,800 dynamic conditions per configuration, compare execution strategies and action horizons with fixed checkpoints, and fine-tune three architectures using dynamic and mixed data. Faster motion degrades every policy but not uniformly; execution timing and horizon can change the behavior of the same policy; and asynchronous replanning benefits from frequent feedback only when the action queue avoids underflow. Dynamic demonstrations improve each architecture under at least one recipe, with gains varying by temporal conditioning, data composition, and training budget. A small-scale physical study shows task-dependent execution-strategy differences consistent with simulation. LIBERO-Dynamic makes object motion, inference latency, and action execution explicit and controllable for evaluating and improving robot policies in moving-target manipulation.

from the abstract

Pith says

Machine review · deepseek-v4-flash · 20 Oct 2026 · what was read was not recorded · not checked by a person

It can be wrong. Challenge this review or reply to it.

Holds up. LIBERO-Dynamic is a well-engineered benchmark that fills a clear gap; limitations are openly discussed, no red flags, deserves acceptance with minor revisions. This line is wrong

5 minor

How often objections from reviews like this one are right: not yet measured

Are you an author? What you can do.

Not checked: figures, code, what the 54 references say.

5 minor objections

The paper says, at §3.4, Latency-Aware Execution Protocols

Pith says: minor objection 1 of 5

The description of the asynchronous queue blending weights (0.3 old, 0.7 new) and the prefetch trigger ratio (0.5K) would benefit from a brief justification or citation to show that these values are standard or effective. The current text presents them as evaluation settings without motivation.

The paper says, at §4.1, Cross-Model Evaluation under Dynamics

Pith says: minor objection 2 of 5

Table 2 reports a 'Ret.' column that normalizes dynamic All SR by the speed=0 result. The speed=0 result is obtained from a separate evaluation (presumably the static baseline). Clarify whether the speed=0 evaluation uses the same episodes with zero object motion, and whether that evaluation is subject to the same latency calibration. The caption says 'Static baseline' but the table header says 'speed= 0'; please ensure consistent terminology.

show 3 more

The paper says, at §4.3, Training with Dynamic Data

Pith says: minor objection 3 of 5

The fine-tuning experiments for π0.5 use different training steps for dynamic-only (1,500) and mixed (3,000) recipes. The paper notes this and states that the comparison is descriptive. However, it would be helpful to include a controlled comparison where both recipes use the same number of steps (e.g., both at 3,000) to better isolate the effect of data composition. This is not a requirement but could strengthen the analysis.

The paper says, at Figure 4

Pith says: minor objection 4 of 5

In Figure 4, the zero-latency synchronous curve (panel a) is labeled as 'τ = 0' but it actually represents an idealized diagnostic where inference latency is set to zero. The caption says 'τ = 0', which is clear, but the legend line says 'τ = 0' without further explanation. Consider adding a footnote that this is a diagnostic condition, not a realistic deployment setting.

The paper says, at §A.3, Dynamic-Demonstration Generation Details

Pith says: minor objection 5 of 5

The retry cascade logic is described in detail, but the number of fallback attempts per condition is not reported. A brief summary of how many conditions required fallbacks and how many were ultimately excluded would help readers assess the coverage of the generated dataset.

What would settle it

To further validate the platform, one could compare key results (e.g., Table 2 model ranking or Figure 4 Async sweep) when latency is drawn from the empirical distribution (e.g., random sample per inference) instead of fixed median. If the relative model ordering or qualitative findings (e.g., underflow threshold) change significantly, this would quantify the sensitivity and guide future benchmark design.

from the skeptic's reading

How the review went

4 readings by deepseek-v4-flash · 20 Oct 2026

  1. Reader holds up confidence high
  2. Skeptic agrees 32 s · 19,466 tokens read, 2,997 written
  3. Editor wrote the verdict above 13 s · 20,211 tokens read, 1,064 written
  4. Referee accept: 5 minor 20 s · 19,742 tokens read, 1,785 written

Not run on this paper: the simulated authors' defence.

Pith did not check this review's objections against the paper.

The review in full

The review in brief, machine-written

The key thing to know: this paper extends LIBERO with controlled object motion and latency-aware execution, producing a reproducible benchmark for dynamic manipulation. It is thorough, honest about its limitations, and should be accepted.

What is new: LIBERO-Dynamic adds three motion families and three speed tiers to all 40 LIBERO tasks, with model-specific inference latency calibrated and baked into the simulation loop. It compares six pretrained policies, multiple execution strategies (sync, async, RTC, VLASH), and action horizons. It also generates 9,000 dynamic demonstrations from static demos via online motion compensation and shows that fine-tuning on dynamic data can improve performance, albeit with task- and architecture-dependent gains.

Credit where due: the engineering is solid. The friction-driven contact platform is a plausible way to add motion without breaking the simulator. The latency calibration procedure is documented in detail, and the execution strategy comparisons are controlled (same checkpoint, same episodes). The paper is unusually transparent about its own limitations—it explicitly notes that it does not cover latency variance, that the fine-tuning recipes differ in steps, that best horizons are measured rather than validated, and that real-robot validation is small-scale. This honesty is a strength, not a weakness.

Soft spots: the incremental novelty is real. DOMINO, DynamicVLA/DOM, and ReflexBench already cover dynamic manipulation with latency awareness. LIBERO-Dynamic’s contribution is the scale (40 tasks), the systematic comparison of execution strategies, and the data mixture study. The static-to-dynamic conversion is hand-designed and condition-matched, so the fine-tuning results primarily measure adaptation to known dynamics, not generalization to unseen motion. The real-robot validation covers only four conditions on one platform and is consistent with simulation rather than providing fundamentally new evidence.

None of these are load-bearing flaws. The paper is careful and the claims are appropriately bounded.

Who benefits: robotics researchers who need a standardized testbed for evaluating policies on moving targets, especially those studying inference latency and execution timing. The benchmark code and data are released.

Verdict: This paper deserves a serious referee and should be accepted. I would support it after minor revisions—perhaps adding a brief discussion of how the results might change if latency variance were modeled, or a clearer separation of the fine-tuning step counts in the main text. But the paper is ready as is.

The strongest objection

After careful review, I do not identify a load-bearing concern that undermines the central claim. The paper's construction is methodical, the benchmark dimensions are clearly specified, and the limitations are explicitly stated (b142–b145, b307, b308, b321, b327). The weakest assumption—using median-calibrated latency—is acknowledged by the authors, and the benchmark is designed for reproducibility rather than exhaustive realism. The static-to-dynamic demonstration conversion is hand-designed but internally consistent for the benchmark's goals. Fine-tuning comparisons are appropriately caveated. The paper supports its claim of providing a reproducible platform for evaluating and improving policies in dynamic manipulation.

The reader's weakest assumption (median-calibrated latency) is correctly identified and acknowledged by the authors. This concern is real but not load-bearing for the paper's central claim, because the benchmark is designed as a controlled, reproducible testbed. The paper's experiments are internally consistent, limitations are transparent, and the code and data are released. No hidden contradiction or unsupported inference was found. The reader's ACCEPT verdict with HIGH confidence is appropriate.

The referee's report

If the claims hold, LIBERO-Dynamic fills a clear gap in existing manipulation benchmarks by making object motion, inference latency, and execution timing explicit and controllable. This is important because most simulation benchmarks treat policy inference as instantaneous and targets as static, obscuring a critical temporal dimension for deployment. The benchmark's controlled setup (40 tasks, three motion families, three speeds, latency-calibrated protocols) enables systematic comparisons that were previously difficult. The paper is transparent about limitations (oracle horizon selection, median latency, non-isolated data composition in fine-tuning, small-scale real-robot validation), which strengthens trust. The construction of 9,000 dynamic demonstrations and the open-source release of code and configurations further increase the benchmark's utility. The cross-model evaluation, execution strategy analysis, and fine-tuning experiments provide concrete empirical findings (e.g., speed sensitivity differs across models, queue underflow limits async gains, temporal conditioning interacts with dynamic data) that are valuable for the community.

Pith's reading of the paper (a machine's summary, not the authors')

The paper establishes LIBERO-Dynamic, an extension of all 40 LIBERO manipulation tasks into dynamic settings where target objects move continuously during robot operation and the simulation world advances according to each model's measured inference latency. The benchmark defines three motion families (linear, curvilinear, zigzag) and three speed tiers, and includes four execution protocols (synchronous hold, asynchronous queue, real-time chunking, future-state-aware streaming). The core finding is that dynamic manipulation success depends on both the policy architecture and the timing of inference relative to execution: faster motion degrades every evaluated model, but the degree varies; the same policy can behave very differently under different execution horizons and queue strategies; and dynamic-data fine-tuning improves performance for all architectures tested, though gains depend on temporal conditioning and training budget. The work also provides a dataset of 9,000 quality-filtered dynamic demonstrations and a small-scale real-robot validation that reproduces timing-related effects observed in simulation.

What carries the argument. The key machinery is a unified simulation loop that couples controlled object motion (driven through MuJoCo contact friction by an invisible platform, never by teleportation) with model-specific, pre-calibrated inference latency. The physical world continues to advance during inference; the number of steps the world advances equals the ceiling of τ×f_c (f_c=20 Hz). The same loop supports both evaluation and training: a static-to-dynamic demonstration converter replays verified static trajectories with online target-relative motion compensation, producing 9,000 filtered dynamic trajectories across 40 tasks, 3 trajectory families, 3 speeds, and 5 motion phases. Four execution protocols (Sync, Async, RTC, VLASH) make the timing of observation, inference, and action explicit and separable.

Core claim. The central discovery is that controllable, latency-aware dynamic manipulation benchmarks reveal coupled dependencies between object speed, execution strategy, and action horizon that are hidden in static benchmarks. Six pretrained policies evaluated on 1,800 shared dynamic episodes show that raising object speed from 0.03 to 0.09 m/s lowers success rates by 4.7–16.3 percentage points with model-dependent declines. With a fixed checkpoint, switching between synchronous hold, asynchronous queue, and real-time chunking changes the best-measured success rate from 37.8% to 48.4% on the same tasks; asynchronous replanning only benefits from frequent feedback when the action queue is sustained without underflow. Dynamic fine-tuning of three architectures (π0.5, FastWAM-Joint, and a history-augmented Joint variant) yields gains of up to 21.7 percentage points, but the magnitude depends on the form of temporal conditioning, data composition, and training step count.

The premise it rests on. The benchmark calibrates each model's inference latency once as a median of many measurements and then freezes it; it does not account for runtime latency variance, and if such variance qualitatively changes execution (e.g., by causing intermittent queue underflow), the quantitative rankings among models and strategies may not generalize to real deployment where latency fluctuates.

If the paper is right (Pith's inferences, not the authors')

Robot policy evaluation that treats inference as instantaneous or keeps objects static risks overestimating real-world performance; future benchmarks should adopt latency-aware, motion-controlled protocols.

The finding that asynchronous action queues require a minimum horizon to avoid underflow provides a concrete design rule for deploying chunked-action policies in dynamic settings: the horizon must exceed the ceiling of (latency × control rate).

Dynamic fine-tuning improves every tested architecture, but gains vary with temporal conditioning; this motivates architecture-level design for explicit temporal encoding (e.g., history frames, future-state prediction) rather than generic static-to-dynamic fine-tuning.

Execution strategy and action horizon alter the behavior of a fixed checkpoint by up to 10+ percentage points, implying that deployment-time tuning of these parameters is at least as important as model selection for dynamic tasks.

The 1,800-condition benchmark and 9,000-demonstration dataset provide a shared testbed for comparing dynamic manipulation methods, enabling controlled ablations of motion speed, latency, and execution protocol that were previously scattered across different task sets.

If the benchmark were extended to include runtime latency jitter (currently acknowledged as a limitation), the observed ordering among execution strategies might change; asynchronous methods that are robust to large jitter could pull ahead of synchronous ones beyond the margins seen here.

The trajectory families (linear, curvilinear, zigzag) differ in direction continuity but do not form a monotonic difficulty scale; future work could design trajectory families that explicitly test predictive tracking vs. reactive correction by controlling acceleration discontinuities.

The static-to-dynamic demonstration pipeline, while tuned for LIBERO, offers a general template for upgrading any static manipulation dataset to dynamic conditions, provided the task has a repeatable reference trajectory; this could multiply the utility of existing large static datasets like DROID or Open X-Embodiment for dynamic scenarios.

The real-robot validation covers only four tasks on one platform; its main value is as a sanity check, but the paper's central claims about timing-dependent performance would be substantially strengthened by a larger-scale real-robot replication, especially under latency jitter.

Check this review

Pith stored no record of this review's passes and took no fingerprint of it. Reviews made by code from 13 Nov 2026 on carry both.

How often reviews like this one are right

Reviews like this one: written by deepseek-v4-flash, where what was read was not recorded.

How often objections from reviews like this one are right: not yet measured

Every value above, in the review's public record (JSON)

Reviews by people

This review was written by a machine. No person has reviewed this paper on Pith yet. A scientist listed on Pith can review it for a fee, and their review is published here whatever it concludes.

Request a human review

How Pith checks papers · The full audit of this paper

The paper

Abstract

Most simulation benchmarks for robotic manipulation keep target objects static and treat policy inference as instantaneous. These assumptions simplify evaluation but hide how a moving target evolves during computation and how stale an observation is when an action is executed. We introduce LIBERO-Dynamic, a dynamic extension of all 40 LIBERO tasks. It drives objects through contact friction along three trajectory families and speed tiers, while advancing the physical world according to calibrated, model-specific inference latency. The same mechanism supports training by replaying verified static demonstrations and compensating robot actions online for target displacement, producing 9,000 quality-filtered dynamic trajectories. We evaluate six pretrained policies on 1,800 dynamic conditions per configuration, compare execution strategies and action horizons with fixed checkpoints, and fine-tune three architectures using dynamic and mixed data. Faster motion degrades every policy but not uniformly; execution timing and horizon can change the behavior of the same policy; and asynchronous replanning benefits from frequent feedback only when the action queue avoids underflow. Dynamic demonstrations improve each architecture under at least one recipe, with gains varying by temporal conditioning, data composition, and training budget. A small-scale physical study shows task-dependent execution-strategy differences consistent with simulation. LIBERO-Dynamic makes object motion, inference latency, and action execution explicit and controllable for evaluating and improving robot policies in moving-target manipulation.

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. No sentence in this record was written by a model.

abstract compiled · 2026-10-20

The authors' main claim, as the classifier picked it out 0.64

“LIBERO-Dynamic makes object motion, inference latency, and action execution explicit and controllable for evaluating and improving robot policies in moving-target manipulation.”

How the main claim is stated: asserted 1.00

What the abstract itself states, as probabilities: claims first 0.09 · states limitation 0.44 · human participants 0.03 · controlled comparison 0.84 · randomized assignment 0.02 · states quantitative result 0.94 · compares with prior results 0.79 · mentions code or data availability 0.06

The abstract, sentence by sentence

This table labels the abstract only (arXiv listing abstract v1). The number is the probability the classifier gave the label.

SentenceRoleThe authors' own workStates the main conclusionStrength
Most simulation benchmarks for robotic manipulation keep target objects static and treat policy inference as instantaneous. background 0.96 0.04 0.02 asserted 1.00
These assumptions simplify evaluation but hide how a moving target evolves during computation and how stale an observation is when an action is executed. gap 0.85 0.07 0.03 asserted 0.99
We introduce LIBERO-Dynamic, a dynamic extension of all 40 LIBERO tasks. contribution 1.00 0.94 0.04 asserted 1.00
It drives objects through contact friction along three trajectory families and speed tiers, while advancing the physical world according to calibrated, model-specific inference latency. method 0.93 0.93 0.04 asserted 1.00
The same mechanism supports training by replaying verified static demonstrations and compensating robot actions online for target displacement, producing 9,000 quality-filtered dynamic trajectories. method 0.98 0.95 0.05 asserted 1.00
We evaluate six pretrained policies on 1,800 dynamic conditions per configuration, compare execution strategies and action horizons with fixed checkpoints, and fine-tune three architectures using dynamic and mixed data. method 0.84 0.97 0.04 asserted 0.93
Faster motion degrades every policy but not uniformly; execution timing and horizon can change the behavior of the same policy; and asynchronous replanning benefits from frequent feedback only when the action queue avoids underflow. result 0.97 0.94 0.30 asserted 1.00
Dynamic demonstrations improve each architecture under at least one recipe, with gains varying by temporal conditioning, data composition, and training budget. result 1.00 0.94 0.18 asserted 0.99
A small-scale physical study shows task-dependent execution-strategy differences consistent with simulation. result 0.85 0.94 0.05 asserted 0.67
LIBERO-Dynamic makes object motion, inference latency, and action execution explicit and controllable for evaluating and improving robot policies in moving-target manipulation. contribution 0.91 0.84 0.67 asserted 1.00

What kind of paper the classifier reports this is: dataset or benchmark 1.00 · evidence from simulation or computation 0.62

What was examined, and what was not

Sources read

  • arXiv listing abstract v1

Not examined

  • full text
  • figures and tables
  • references and citing sentences
  • relations to other papers

Levels run

  • P abstract

Not run

  • S sections
  • T body sentences
  • C citations
  • R cross-paper relations

How the labels were made

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-10-20: 1 request, 7,519 input tokens.

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. This record as JSON

Figures (11)
Figure 1 of the paper
Figure 1. Overview of LIBERO-Dynamic: (a) dynamic real-world manipulation and conventional simulation evaluation; (b) latency-aware evaluation; and (c) comparisons across execution and training settings. continue to advance during inference, leaving observations and actions ideally aligned in time. In a static scene, waiting for inference does not change the state of the target object; when the object continues to move, however, the observation underlying a returned action may already be stale. DOMINO (Fang et al., 2026) studies moving-object interaction, while DOM (Xie et al., 2026) and…
Figure 2 of the paper
Figure 2. Benchmark dimensions, dynamic-data construction, execution protocols, and representative
Figure 3 of the paper
Figure 3. Normalized end-effector position density in the robot base frame for original
Figure 4 of the paper
Figure 4. π0.5 (PyTorch) SR versus action horizon under (a) Sync, (b) Async, (c) RTC, and (d) VLASH. Panel (a) includes a zero-latency diagnostic; panel (b) overlays steady underflow (right axis)
Figure 5 of the paper
Figure 5. Dynamic-data fine-tuning under asynchronous execution.
Figure 6 of the paper
Figure 6. Real-robot task sequences. Each panel shows three keyframes from one successful rollout;
Figure 7 of the paper
Figure 7. Real-robot results at K ∈ {4, 8}. The manipulation tasks report successful trials out of 50; whack-a-mole reports successful presses over 50 × 30 s sessions. 5 Limitations LIBERO-Dynamic builds on LIBERO tasks and scenes, using prescribed trajectories and fixed calibrated latency to enable repeatable dynamic evaluation. It does not yet cover deployment-time latency fluctuations or more complex, unpredictable object motion. The fine-tuning study separates training and evaluation by motion phase while sharing tasks, paths, and speeds; it therefore primarily evaluates adaptation to…
Figure 8 of the paper
Figure 8. Dynamic demonstrations for four representative tasks under the three motion families. In
Figure 9 of the paper
Figure 9. reports the complete synchronous and asynchronous sweeps underlying the best-measured operating points in
Figure 10 of the paper
Figure 10. Extended π0.5 (PyTorch) strategy sweeps, including two RTC settings and asynchronous underflow. Markers are measured results; smooth SR curves are visual guides. A.9 Best-Measured Execution-Strategy Results
Figure 11 of the paper
Figure 11. Conveyor handover: seven keyframes from one successful episode.

Figures are the authors', reproduced from the paper.

References (54)
  1. [1] Pranav Atreya, Karl Pertsch, Tony Lee, Moo Jin Kim, Arhan Jain, Artur Kuramshin, Clemens Eppner, et al. RoboArena : Distributed real-world evaluation of generalist robot policies. arXiv preprint arXiv:2506.18123, 2025 on PitharXiv
  2. [5] Anthony Brohan, Noah Brown, Justice Carbajal, Yevgen Chebotar, Xi Chen, Krzysztof Choromanski, et al. RT-2 : Vision-language-action models transfer web knowledge to robotic control. arXiv preprint arXiv:2307.15818, 2023 on PithPith's reviewarXiv
  3. [7] Edwin Catmull and Raphael Rom. A class of local interpolating splines. In Robert E. Barnhill and Richard F. Riesenfeld, editors, Computer Aided Geometric Design, pages 317--326. Academic Press, 1974. doi:10.1016/B978-0-12-079050-0.50020-5 on Pithdoi:10.1016/b978-0-12-079050-0.50020-5
  4. [8] Tianxing Chen et al. RoboDojo : A unified sim-and-real benchmark for comprehensive evaluation of generalist robot manipulation policies. arXiv preprint arXiv:2607.04434, 2026 a on PithPith's reviewarXiv
  5. [10] Rongxu Cui, Zongzheng Zhang, Jingrui Pang, Haohan Chi, Jinbang Guo, Saining Zhang, Shaoxuan Xie, Xin Jin, Yao Mu, Jiaolong Yang, Guocai Yao, Xianyuan Zhan, Ya-Qin Zhang, and Hao Zhao. LIBERO-Safety : A comprehensive benchmark for physical and semantic safety in vision-language... on PithPith's reviewarXiv
  6. [12] Senyu Fei, Siyin Wang, Junhao Shi, Zihao Dai, Jikun Cai, Pengfang Qian, Li Ji, et al. LIBERO-Plus : In-depth robustness analysis of vision-language-action models. arXiv preprint arXiv:2510.13626, 2025 on PithPith's reviewarXiv
  7. [13] Stephen James, Zicong Ma, David Rovick Arrojo , and Andrew J. Davison. RLBench : The robot learning benchmark & learning environment. IEEE Robotics and Automation Letters, 5 0 (2): 0 3019--3026, 2020 on Pith
  8. [14] Alexander Khazatsky, Karl Pertsch, Suraj Nair, Ashwin Balakrishna, Sudeep Dasari, Siddharth Karamcheti, et al. DROID : A large-scale in-the-wild robot manipulation dataset. In Robotics: Science and Systems (RSS), 2024 on Pith
  9. [15] Moo Jin Kim, Karl Pertsch, Siddharth Karamcheti, Ted Xiao, Ashwin Balakrishna, Suraj Nair, Rafael Rafailov, Ethan Foster, Grace Lam, Pannag Sanketi, et al. OpenVLA : An open-source vision-language-action model. arXiv preprint arXiv:2406.09246, 2024 on PithPith's reviewarXiv
  10. [16] Moo Jin Kim, Chelsea Finn, and Percy Liang. Fine-tuning vision-language-action models: Optimizing speed and success. In Robotics: Science and Systems (RSS), 2025. arXiv:2502.19645 on PithPith's reviewarXiv
  11. [17] Bo Liu, Yifeng Zhu, Chongkai Gao, Yihao Feng, Qiang Liu, Yuke Zhu, and Peter Stone. LIBERO : Benchmarking knowledge transfer for lifelong robot learning. In Advances in Neural Information Processing Systems, 2023 on Pith
  12. [18] Yuxiang Lu, Zhe Liu, Xianzhe Fan, Zhenya Yang, Jinghua Hou, Junyi Li, Kaixin Ding, and Hengshuang Zhao. FASTER : Rethinking real-time flow VLA s. arXiv preprint arXiv:2603.19199, 2026. URL https://arxiv.org/abs/2603.19199 on PithPith's reviewarXiv
  13. [19] Yao Mu et al. RoboTwin : Dual-arm robot benchmark with generative digital twins. arXiv preprint arXiv:2409.02920, 2024 on PithPith's reviewarXiv
  14. [22] Kohei Sendai, Maxime Alvarez, Tatsuya Matsushima, Yutaka Matsuo, and Yusuke Iwasawa. Leave no observation behind: Real-time correction for VLA action chunks. arXiv preprint arXiv:2509.23224, 2025. URL https://arxiv.org/abs/2509.23224 on PitharXiv
  15. [23] Jingwen Sun, Wenyao Zhang, Zekun Qi, Shaojie Ren, Zezhi Liu, Hanxin Zhu, Guangzhong Sun, Xin Jin, and Zhibo Chen. VLA-JEPA : Enhancing vision-language-action model with latent world model. arXiv preprint arXiv:2602.10098, 2026 on PitharXiv
  16. [24] Jiaming Tang, Yufei Sun, Yilong Zhao, Shang Yang, Yujun Lin, Zhuoyang Zhang, James Hou, Yao Lu, Zhijian Liu, and Song Han. VLASH : Real-time VLA s via future-state-aware asynchronous inference. arXiv preprint arXiv:2512.01031, 2025. URL https://arxiv.org/abs/2512.01031 on PithPith's reviewarXiv
  17. [25] 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, pages 5026--5033, 2012. doi:10.1109/IROS.2012.6386109 on Pithdoi:10.1109/iros.2012.6386109
  18. [26] Guodong Wang, Chenkai Zhang, Qingjie Liu, Jinjin Zhang, Jiancheng Cai, Junjie Liu, and Xinmin Liu. LIBERO-X : Robustness litmus for vision-language-action models. arXiv preprint arXiv:2602.06556, 2026 on PitharXiv
  19. [27] Haozhe Xie et al. DynamicVLA : A vision-language-action model for dynamic object manipulation. arXiv preprint arXiv:2601.22153v1, 2026. URL https://arxiv.org/abs/2601.22153v1 on PithPith's reviewarXiv
  20. [28] Adina Yakefu et al. RoboChallenge : Large-scale real-robot evaluation of embodied policies. arXiv preprint arXiv:2510.17950, 2025 on PitharXiv
show 34 more
  1. [29] Tianyuan Yuan, Zibin Dong, Yicheng Liu, and Hang Zhao. Fast-WAM : Do world action models need test-time future imagination? arXiv preprint arXiv:2603.16666, 2026 on PithPith's reviewarXiv
  2. [30] Borong Zhang, Jiahao Li, Jiachen Shen, Yuhao Zhang, Yishuai Cai, Yuanpei Chen, Juntao Dai, Jiaming Ji, and Yaodong Yang. VLA-Arena : An open-source framework for benchmarking vision-language-action models. In Proceedings of the 43rd International Conference on Machine Learning... on Pith
  3. [31] Xueyang Zhou, Yangming Xu, Guiyao Tie, Yongchao Chen, Guowen Zhang, Duanfeng Chu, Pan Zhou, and Lichao Sun. LIBERO-PRO : Towards robust and fair evaluation of vision-language-action models beyond memorization. arXiv preprint arXiv:2510.03827, 2025 on PithPith's reviewarXiv
  4. [32] Brohan, Anthony and Brown, Noah and Carbajal, Justice and Chebotar, Yevgen and Chen, Xi and Choromanski, Krzysztof and others , journal = on Pith
  5. [33] Octo: An Open-Source Generalist Robot Policy , author =. arXiv preprint arXiv:2405.12213 , year = on PithPith's reviewarXiv
  6. [34] Kim, Moo Jin and Pertsch, Karl and Karamcheti, Siddharth and Xiao, Ted and Balakrishna, Ashwin and Nair, Suraj and Rafailov, Rafael and Foster, Ethan and Lam, Grace and Sanketi, Pannag and others , journal = on Pith
  7. [35] Atreya, Pranav and Pertsch, Karl and Lee, Tony and Kim, Moo Jin and Jain, Arhan and Kuramshin, Artur and Eppner, Clemens and others , journal = on Pith
  8. [36] Yakefu, Adina and others , journal = on Pith
  9. [37] Chen, Tianxing and others , journal = on Pith
  10. [38] Liu, Bo and Zhu, Yifeng and Gao, Chongkai and Feng, Yihao and Liu, Qiang and Zhu, Yuke and Stone, Peter , booktitle = on Pith
  11. [39] James, Stephen and Ma, Zicong and. IEEE Robotics and Automation Letters , volume = on Pith
  12. [40] Mu, Yao and others , journal = on Pith
  13. [41] Fei, Senyu and Wang, Siyin and Shi, Junhao and Dai, Zihao and Cai, Jikun and Qian, Pengfang and Ji, Li and others , journal = on Pith
  14. [42] Zhou, Xueyang and Xu, Yangming and Tie, Guiyao and Chen, Yongchao and Zhang, Guowen and Chu, Duanfeng and Zhou, Pan and Sun, Lichao , journal = on Pith
  15. [43] Wang, Guodong and Zhang, Chenkai and Liu, Qingjie and Zhang, Jinjin and Cai, Jiancheng and Liu, Junjie and Liu, Xinmin , journal = on Pith
  16. [44] Zhang, Borong and Li, Jiahao and Shen, Jiachen and Zhang, Yuhao and Cai, Yishuai and Chen, Yuanpei and Dai, Juntao and Ji, Jiaming and Yang, Yaodong , booktitle = on Pith
  17. [45] Cui, Rongxu and Zhang, Zongzheng and Pang, Jingrui and Chi, Haohan and Guo, Jinbang and Zhang, Saining and Xie, Shaoxuan and Jin, Xin and Mu, Yao and Yang, Jiaolong and Yao, Guocai and Zhan, Xianyuan and Zhang, Ya-Qin and Zhao, Hao , booktitle =. 2026 , note = on Pith
  18. [46] Towards Generalizable Robotic Manipulation in Dynamic Environments , author =. arXiv preprint arXiv:2603.15620 , year = on PithPith's reviewarXiv
  19. [47] Xie, Haozhe and others , journal =. 2026 , url = on Pith
  20. [48] Reflex: Enabling Fast and Predictive Vision-Language-Action Models for Reaction-Critical Manipulation , author =. arXiv preprint arXiv:2608.14379 , year = on PithPith's reviewarXiv
  21. [49] Tang, Jiaming and Sun, Yufei and Zhao, Yilong and Yang, Shang and Lin, Yujun and Zhang, Zhuoyang and Hou, James and Lu, Yao and Liu, Zhijian and Han, Song , journal =. 2025 , eprint = on Pith
  22. [50] Real-Time Execution of Action Chunking Flow Policies , author =. arXiv preprint arXiv:2506.07339 , year =. 2506.07339 , archivePrefix = on PithPith's reviewarXiv
  23. [51] Training-Time Action Conditioning for Efficient Real-Time Chunking , author =. arXiv preprint arXiv:2512.05964 , year =. 2512.05964 , archivePrefix = on PitharXiv
  24. [52] Sendai, Kohei and Alvarez, Maxime and Matsushima, Tatsuya and Matsuo, Yutaka and Iwasawa, Yusuke , journal =. Leave No Observation Behind: Real-Time Correction for. 2025 , eprint = on Pith
  25. [53] Lu, Yuxiang and Liu, Zhe and Fan, Xianzhe and Yang, Zhenya and Hou, Jinghua and Li, Junyi and Ding, Kaixin and Zhao, Hengshuang , journal =. 2026 , eprint = on Pith
  26. [54] Sun, Jingwen and Zhang, Wenyao and Qi, Zekun and Ren, Shaojie and Liu, Zezhi and Zhu, Hanxin and Sun, Guangzhong and Jin, Xin and Chen, Zhibo , journal = on Pith
  27. [55] Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success , author =. Robotics: Science and Systems (RSS) , year = on Pith
  28. [56] Yuan, Tianyuan and Dong, Zibin and Liu, Yicheng and Zhao, Hang , journal = on Pith
  29. [57] _0 : A Vision-Language-Action Flow Model for General Robot Control , author =. arXiv preprint arXiv:2410.24164 , year = on PithPith's reviewarXiv
  30. [58] _. arXiv preprint arXiv:2504.16054 , year = on PithPith's reviewarXiv
  31. [59] Cadene, Remi and Aliberts, Simon and Capuano, Francesco and Aractingi, Michel and Zouitine, Adil and Kooijmans, Pepijn and Choghari, Jade and Russi, Martino and Pascal, Caroline and Palma, Steven and Shukor, Mustafa and Moss, Jess and Soare, Alexander and Aubakirova, Dana and ... on PitharXiv
  32. [60] Todorov, Emanuel and Erez, Tom and Tassa, Yuval , booktitle =. 2012 , doi = on Pith
  33. [61] A Class of Local Interpolating Splines , author =. Computer Aided Geometric Design , editor =. 1974 , doi = on Pith
  34. [62] Khazatsky, Alexander and Pertsch, Karl and Nair, Suraj and Balakrishna, Ashwin and Dasari, Sudeep and Karamcheti, Siddharth and others , booktitle = on Pith
Cited by 1 Pith paper

Challenges and discussion

Disagree with the review, or know something it missed? Leave a comment, or challenge this review formally: a challenge becomes a permanent entry in this paper's record, and its decision is published here.

No comments yet.

Sign in to comment