Robotics · arXiv:2610.05734 v1 · submitted 5 Oct 2026 · 29 references
FreeSpeed: Training-Free Speed Control for Generative Robot Policies
The paper says
“Online control of execution speed is essential for deploying robot policies in real-world scenarios, as robots may need to speed up under time constraints or slow down to facilitate human interaction and improve safety.”
more of the abstract
However, imitation-learned policies inherit the execution speed of their demonstrations, and test-time speed modification can introduce unrecoverable out-of-distribution observations, reducing task success. We observe that the directional inconsistency of action chunks reflects task-phase criticality, indicating how aggressively action step lengths can be modified while preserving task success. Based on this observation, we introduce FreeSpeed, a training-free module that post-processes action chunks from pretrained policies. FreeSpeed resamples each predicted chunk at the requested rate, then uses directional inconsistency between adjacent actions as the primary signal for rescaling. This signal adaptively determines how closely the execution speed can approach the requested speed, allowing flexible speed adjustment within the evaluated limits without compromising task success. Across three policy families and 50 simulated tasks, FreeSpeed supports online speed changes, with realized execution rates spanning 0.22x to 2.53x among settings that preserve per-task success. Across four real-world manipulation tasks, FreeSpeed achieves an average success rate of 94.0%, matching the frozen policy's 93.8%, while realizing execution rates from 0.38x to 1.97x.
from the abstract
Pith says
Machine review · deepseek-v4-flash · 7 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, with conditions. FreeSpeed is a credible training-free speed controller, with a mechanistic claim thinner than the empirical results. This line is wrong
4 major objections · 4 minor
How often objections from reviews like this one are right: not yet measured
Not checked: figures, code, what the 29 references say.
The paper says, at 5.2, Eqs. 2-3
The review did not quote this passage.
Pith says: Objection 1 of 4 · major
The central mechanistic claim, that I_cos tracks task-phase criticality in predicted chunks, is supported by Figure 4a computed on only five LIBERO-90 human demonstration traces, and the observation-distribution mechanism in Section 5.5 is measured on a single task (Task 60).
more of objection 1
The bridging statement in Section 5.2, that the policy 'learns these features and reproduces them in its predicted action chunks', is asserted rather than demonstrated across tasks and policy families. Please plot I_cos over predicted chunks (rather than demonstrations) for all three policy architectures across more tasks, showing peak alignment with grasp and release phases, and provide evidence that the max aggregation in Eq. 3 is the appropriate statistic rather than, say, the mean directional inconsistency.
The paper says, at 5.5, Appendix E.2
The review did not quote this passage.
Pith says: Objection 2 of 4 · major
The in-distribution advantage of FreeSpeed is partly confounded by systematically less aggressive execution: at the 2x command, w/o CR scales every retained action by 2.0 while FreeSpeed caps at f_s = 0.9*rho = 1.8 and reverts to f_c = 1 on high-CI chunks, giving a realized rate of 1.34x versus 1.55x for vanilla on pi0.5 LIBERO (Section D.6).
more of objection 2
The Appendix E.2 constant-factor sweep is the right control and supports the claim on tasks 34, 48, and 86, but the coverage is narrow. Please add matched-realized-rate comparisons, for example FreeSpeed at 2x against w/o CR at a rate-matched command, across a larger portion of the task set, or extend the E.2 sweep, so that success is attributed to CI gating rather than to overall reduced aggression.
The paper says, at 4.2, Appendix C.2, Table A4
The review did not quote this passage.
Pith says: Objection 3 of 4 · major
Equation (5) contains hand-set constants, lambda = 50, f_s = 0.9*rho, the f_c endpoints in Table A4, and the expanded controller bound L = 4 in Appendix C.3, whose selection provenance is not stated. The sensitivity analysis in Figure 7 shows robustness to lambda and f_s over a range, which mitigates the concern, but the paper should state explicitly whether …
more of objection 3
… these values were fixed before or after observing the evaluation tasks, for example whether they were selected on a held-out task subset. If they were tuned on the reported tasks, the claim that the schedule is training-free and transferable is weakened and should be reworded.
The paper says, at Tables 2, 3, A8
The review did not quote this passage.
Pith says: Objection 4 of 4 · major
Success rates are reported as point estimates from 50 simulation and 20 real-robot rollouts per condition without confidence intervals. Several per-task claims that the method matches or exceeds the reference rest on small samples (Table A8's 92 vs 92 at n = 50 has wide binomial uncertainty, and Table 3's per-task real-robot entries use 20 trials each). Please report binomial confidence intervals or tests at least at the task level for the claims that success is preserved.
4 minor objections
The paper says, at 5.2, Figure 4a
Pith says: minor objection 1 of 4
Figure 4a is computed on human demonstrations rather than on predicted chunks; add a sentence in the text or caption making this explicit, alongside the requested predicted-chunk analysis, since the gating in Eq. 5 operates on predicted chunks.
The paper says, at Figure 5
Pith says: minor objection 2 of 4
The rate axes are inverted on the slowdown panels; annotate the inversion direction clearly, as the current layout is easy to misread.
show 2 more
The paper says, at Table 2
Pith says: minor objection 3 of 4
At the 0.3x command the realized FreeSpeed rate (0.402) exceeds the commanded stride; a one-line note explaining that the conservative deceleration endpoints (f_c > rho in Table A4) cause this would prevent confusion.
The paper says, at 5.7, Figure 7a
Pith says: minor objection 4 of 4
The comparison with SuP and TempoVLA uses success rates reported in other papers under different evaluation settings; appendix D.3 acknowledges this, and a one-line caveat near the figure itself would help readers.
What would settle it
Using the same frozen Flow Matching checkpoints, reproduce Section 5.5's kNN observation-distance analysis (Figures 6b/6c and A5) plus Figure 4a-style phase alignment on three additional LIBERO-90 tasks spanning the paper's own placement groups: task 2 (structure-constrained), task 48 (bounded receptacle), and task 69 (open relational), at the 2x and 0.3x commands with 50 rollouts each. A meaningful pass requires (a) predicted-chunk Icos (not demonstration Icos) to peak at grasp/release for all three tasks, and (b) FreeSpeed's median observation-distance ratio to the 1x reference to stay near the Task 60 value (~1.09-1.13) while w/o CR stays above ~1.6. If either condition fails on any task, the CI-proxy mechanism does not transfer and the claim should be narrowed to 'conservative rescaling preserves success' rather than 'CI-gated rescaling tracks criticality'.
from the skeptic's reading
How the review went
4 readings by deepseek-v4-flash · 7 Oct 2026
- Reader holds up, with conditions
- Skeptic agrees
- Editor wrote the verdict above
- Referee major revision: 4 major objections, 4 minor
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
FreeSpeed works, and the paper is honest about its limits. The training-free speed control is a real capability: frozen action-chunking policies can be slowed or sped up at test time with roughly preserved success across three policy families, 50 simulated tasks, and four real-robot tasks. The comparison is well controlled, and the real-robot numbers are credible. The paper reports its confounds honestly — the 4x realized rate averages only over successful trials, the acceleration ceiling is task-dependent, and deceleration can also fail. That is solid engineering. The soft spot is the mechanistic claim, not the empirical result. CI-as-criticality-proxy rests on five demonstration traces and one OOD measurement task. And there is a genuine confound: at 2x, w/o CR scales every step by 2.0 while FreeSpeed caps at 1.8 and pulls most chunks to fc=1, so FreeSpeed is systematically less aggressive. Part of its in-distribution advantage is conservatism, not CI finding the critical phases. The constant-factor sweep in the appendix partially addresses this, but only on three pick-and-place tasks. Success rates lack confidence intervals, and the constants (lambda=50, fs=0.9rho, fc endpoints) are hand-set; a sensitivity sweep mitigates the circularity, but they are still fitted. None of this breaks the central empirical claim. The open question is transfer: on a task where CI does not peak at the critical phase, the advantage over w/o CR could vanish. The paper deserves a serious referee. I would ask for code, confidence intervals, and at least one task class where CI and criticality do not align. For your own work on test-time control, it is worth citing and worth bringing to reading group.
The strongest objection
The central claim has two layers: (1) empirically, FreeSpeed preserves or exceeds the 1x reference success across 50 simulated and 4 real tasks; (2) mechanistically, it does so because Icos = max_i(1 - cos angle between adjacent delta-p actions) is a policy-independent proxy for phase criticality, so that Equation (5)'s CI-gated rescaling keeps observations in-distribution. Layer (1) is well supported; layer (2), which is the stated novel contribution (§1, §4.1), rests on narrow evidence. Figure 4a shows CI peaks during grasp/placement for only five LIBERO-90 demonstration trajectories, and the observation-distribution mechanism in Section 5.5 (Figures 6, A5, A6) is measured on exactly one task, Task 60. Moreover, Figure 4a measures CI on human demonstrations, whereas the gating operates on predicted chunks; the claim that 'the policy learns these features and reproduces them' (§5.2) is only checked for Task 60. A further confound: at 2x, w/o CR scales every retained action by 2.0 while FreeSpeed caps at fs=0.9rho=1.8 and drops most chunks to fc=1, so FreeSpeed systematically deviates less per replan (realized 1.34x vs 1.55x for vanilla on LIBERO). Figure 6's in-distribution advantage is therefore consistent with the weaker statement that FreeSpeed simply executes less aggressively, not that CI specifically identifies which phases tolerate rescaling. The constant-factor sweep (Appendix E.2) partially answers this, but on three tasks only. None of this refutes the empirical success claim, and the authors honestly state the acceleration ceiling limitation (Appendix F), but the mechanism that distinguishes FreeSpeed from 'rescale conservatively everywhere' is not yet shown to transfer to settings where CI does not peak at critical phases (e.g., tasks with turning free-space motion or oscillatory approaches). If on some task the gating misaligns with criticality, FreeSpeed's advantage over w/o CR could vanish while the frozen-policy claim would still hold it harmless.
The reader's weakest_assumption is that the CI proxy and gating schedule generalize across tasks and policy families, with the OOD mechanism validated on a single task. I agree this is the load-bearing concern, and I find it is genuinely the fragile link: the empirical success-preservation claim is broad (50 sim + 4 real tasks, three policy families), with baseline gaps much larger than 50-rollout noise, but the mechanism—the actual novelty—is supported by five demonstration traces and one OOD task. I added two sharpenings: (a) the criticality claim is made on demonstration trajectories while the gating acts on predicted chunks, a gap only bridged on Task 60; and (b) the 'in-distribution' advantage in Figure 6 is confounded by FreeSpeed's systematically gentler rescaling (fs=0.9rho, fc=1), which is consistent with a weaker 'deviate less everywhere' explanation. The authors' own evidence partially mitigates this: the constant-factor sweep in Appendix E.2 suggests no single constant factor matches FreeSpeed, and the lambda/fs sensitivity plateau in Figures 7b-7c shows the schedule is robust. No ad hominem, no internal-inconsistency claim: the paper is honest about its acceleration ceiling (Appendix F) and its single-task analysis. Because the concern is about the width of the mechanism claim rather than the existence of the effect, CONDITIONAL remains the right verdict; my read does not move it in either direction. The concrete test is a direct replication of Section 5.5's measurement on three tasks spanning the paper's own task-type taxonomy, which would settle transferability in one pass.
The referee's report
The empirical layer of the claim is well supported and the evaluation is unusually well controlled: all methods run on identical frozen checkpoints and initial states, the w/o CR ablation isolates the CI-based rescaling, the Appendix E.2 constant-factor sweep is a genuine control for the 'simply less aggressive' confound, and the sensitivity analysis in Figure 7 explores lambda and f_s. The reproducibility appendix (architectures, controller bounds, hardware, full per-task tables) is complete to an uncommon degree. If the result holds, FreeSpeed is a practical plug-and-play speed-control mechanism with no retraining, user-set and in-episode rates, and both acceleration and deceleration, a capability combination not jointly demonstrated by the compared methods. The risk sits in the mechanistic layer: the CI criticality proxy is validated on five demonstration traces and one out-of-distribution task, and the transferability of the hand-set schedule constants across tasks and policy families is asserted more strongly than the current evidence warrants.
Pith's reading of the paper (a machine's summary, not the authors')
Imitation-learned robot policies inherit a fixed speed from their demonstrations, and the usual route to speed control is expensive retraining tied to task-specific data augmentation. This paper argues that a policy's own predicted action chunk already carries a cheap, training-free signal for how much its speed can be safely modified: the cosine inconsistency between adjacent translation increments, which spikes during critical phases such as grasping and placing. FreeSpeed uses that signal to rescale resampled actions, keeping the frozen policy's observations close to its training distribution while pushing execution toward a user-requested rate. Across three policy families, 50 simulated tasks, and four real-robot tasks, the paper reports that FreeSpeed matches or exceeds the frozen policy's success rate over commands from 0.2x to 4x while realizing execution rates between 0.22x and 2.53x. The contribution is a plug-and-play speed interface for policies that would otherwise require retraining or demonstration reprocessing.
What carries the argument. The central object is the inconsistency-gated rescaling rule ft = fc(rho_t) + (fs(rho_t) - fc(rho_t)) exp(-lambda Icos(A_t)), applied after temporal resampling of the predicted action chunk on a grid of stride rho_t. Here Icos is the maximum of di = 1 - the cosine of the angle between adjacent translation increments within the chunk, which the paper shows peaks during decisive phases like grasping and placement in human demonstrations. The rule does the work: it edges executed step lengths toward the commanded rate when motion is straight and pulls them back toward the 1x reference when directions disagree, so the frozen policy never strays far from its training distribution. A single decay rate lambda = 50 and the endpoint pairs (fs, fc) are shared across tasks and policy families, which is what makes the module plug-and-play rather than task-specific.
Core claim. The paper's central claim is that the maximum cosine inconsistency between consecutive translation increments in a predicted action chunk, Icos = max_i (1 - cos of the angle between adjacent delta-p actions), acts as a training-free proxy for task-phase criticality. Where Icos is small, the chunk is directionally consistent and its speed can be modified aggressively; where Icos is large, the policy is in a decisive phase such as grasping, releasing, or placing, and step lengths should stay close to the original predictions. FreeSpeed operationalizes this by resampling the chunk at the commanded stride and rescaling each action by ft = fc + (fs - fc) exp(-lambda Icos), with a single decay constant lambda = 50 shared across all tasks and policies, fs = 0.9 rho for acceleration, and fc near 1.0 for critical chunks. The paper provides evidence that this adaptive recipe keeps the frozen policy's conditioning observations close to the training distribution, and that it applies to any frozen action-chunking policy without policy updates, additional data, or a learned scheduler.
The premise it rests on. The load-bearing premise is that the maximum cosine inconsistency of a predicted action chunk is a reliable, policy-independent proxy for how much step-length modification a task phase tolerates, so that one fixed schedule (lambda = 50, fs = 0.9 rho) generalizes across tasks and policy families.
If the paper is right (Pith's inferences, not the authors')
A frozen vision-language-action model, world-action model, or flow-matching policy can expose a user-requested speed interface at test time without retraining, finetuning, or demonstration resampling.
Because the rescaling is continuous in Icos, speed can be updated within an episode: a command can slow down for a precision phase and speed up again once motion is straight, with success preserved.
The same single schedule (lambda = 50, fs = 0.9 rho) transfers across three policy families and from simulation to a real humanoid arm, so no per-task tuning is needed.
Realized speedup is capped by how much of the task consists of directionally consistent free motion, a limit the authors acknowledge in Appendix F.
The mechanism suggests a general diagnostic: any action-chunking policy that must run at an off-nominal speed could be instrumented the same way, using chunk geometry rather than learned critics or world models to decide when to trust aggressive step lengths.
If Icos is a faithful phase-criticality signal, it could be combined with the tactile or force feedback the authors mention to gate speed during contact-rich phases, where kinematics alone may not warn of danger.
Because FreeSpeed only rescales and resamples, its ceiling is the frozen policy's own competence: on Robomimic aggressive 4x acceleration still drops success to 55.3%, so the gains are real but conditional on a base policy that succeeds in the first place.
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
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.
The paper
Abstract
Online control of execution speed is essential for deploying robot policies in real-world scenarios, as robots may need to speed up under time constraints or slow down to facilitate human interaction and improve safety. However, imitation-learned policies inherit the execution speed of their demonstrations, and test-time speed modification can introduce unrecoverable out-of-distribution observations, reducing task success. We observe that the directional inconsistency of action chunks reflects task-phase criticality, indicating how aggressively action step lengths can be modified while preserving task success. Based on this observation, we introduce FreeSpeed, a training-free module that post-processes action chunks from pretrained policies. FreeSpeed resamples each predicted chunk at the requested rate, then uses directional inconsistency between adjacent actions as the primary signal for rescaling. This signal adaptively determines how closely the execution speed can approach the requested speed, allowing flexible speed adjustment within the evaluated limits without compromising task success. Across three policy families and 50 simulated tasks, FreeSpeed supports online speed changes, with realized execution rates spanning 0.22x to 2.53x among settings that preserve per-task success. Across four real-world manipulation tasks, FreeSpeed achieves an average success rate of 94.0%, matching the frozen policy's 93.8%, while realizing execution rates from 0.38x to 1.97x.
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 · 2026-10-06
The authors' main claim, as the classifier picked it out 0.55
“Across four real-world manipulation tasks, FreeSpeed achieves an average success rate of 94.0%, matching the frozen policy's 93.8%, while realizing execution rates from 0.38x to 1.97x.”
How the main claim is stated: asserted 1.00
What the abstract itself states, as probabilities: claims first 0.05 · states limitation 0.47 · human participants 0.12 · controlled comparison 0.91 · randomized assignment 0.02 · states quantitative result 0.99 · compares with prior results 0.88 · mentions code or data availability 0.03
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.
| Sentence | Role | The authors' own work | States the main conclusion | Strength |
|---|---|---|---|---|
| Online control of execution speed is essential for deploying robot policies in real-world scenarios, as robots may need to speed up under time constraints or slow down to facilitate human interaction and improve safety. | background 1.00 | 0.06 | 0.03 | asserted 0.66 |
| However, imitation-learned policies inherit the execution speed of their demonstrations, and test-time speed modification can introduce unrecoverable out-of-distribution observations, reducing task success. | background 0.63 | 0.08 | 0.03 | asserted 0.91 |
| We observe that the directional inconsistency of action chunks reflects task-phase criticality, indicating how aggressively action step lengths can be modified while preserving task success. | result 0.50 | 0.89 | 0.10 | asserted 0.58 |
| Based on this observation, we introduce FreeSpeed, a training-free module that post-processes action chunks from pretrained policies. | contribution 0.95 | 0.88 | 0.10 | asserted 1.00 |
| FreeSpeed resamples each predicted chunk at the requested rate, then uses directional inconsistency between adjacent actions as the primary signal for rescaling. | method 0.99 | 0.89 | 0.07 | asserted 1.00 |
| This signal adaptively determines how closely the execution speed can approach the requested speed, allowing flexible speed adjustment within the evaluated limits without compromising task success. | method 0.46 | 0.84 | 0.37 | asserted 0.97 |
| Across three policy families and 50 simulated tasks, FreeSpeed supports online speed changes, with realized execution rates spanning 0.22x to 2.53x among settings that preserve per-task success. | result 1.00 | 0.96 | 0.27 | asserted 1.00 |
| Across four real-world manipulation tasks, FreeSpeed achieves an average success rate of 94.0%, matching the frozen policy's 93.8%, while realizing execution rates from 0.38x to 1.97x. | result 1.00 | 0.97 | 0.30 | asserted 1.00 |
What kind of paper the classifier reports this is: method or tool 1.00 · evidence from observation or experiment 0.92
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, 2026-10-06: 1 request, 6,403 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
References (29)
- [2] Type Task 1× 1.5× 2× 0.5× 0.3× Ref. w/o CR vanilla Free Speed w/o CR vanilla Free Speed w/o CR Free Speed w/o CR Free Speed Structure constrained 2 121 100 98 105 119 93 100 242 194 440 284 32 204 148 160 178 127 125 152 364 327 580 489 40 203 158 176 161 161 149 175 360 338 5... on Pith
- [5] Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. BERT: Pre-training of deep bidirectional transformers for language understanding. InProceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Langua... on Pith
- [8] Qingda Hu, Ziheng Qiu, Jieru Zhao, Zhongxue Gan, and Wenchao Ding. AutoSpeed: Annotation-free stage-adaptive motion speed learning for robot manipulation.arXiv preprint arXiv:2607.01051, 2026a. Yuxuan Hu, Shihao Li, Yang Xiao, Gen Li, Feng Xu, and Jianfei Yang. ActiveVital: Ge... on PithPith's reviewarXiv
- [9] Dong Jing, Jingchen Nie, Tianqi Zhang, Jiaqi Liu, Huaxiu Yao, Zhiwu Lu, and Mingyu Ding. TempoVLA: Learning speed-controllable vision-language-action policies.arXiv preprint arXiv:2606.06491, on PithPith's reviewarXiv
- [10] Ian T. Jolliffe and Jorge Cadima. Principal component analysis: a review and recent developments.Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 374(2065):20150202, on Pith
- [12] Yanshuo Lu, Yuxuan Hu, Shenghai Yuan, Xinyu Zhou, Kuangji Zuo, Bofan Lyu, XiChen Yuan, and Jianfei Yang. EM-Fall: Embodied mmWave sensing for day-and-night fall detection on humanoid robots.arXiv preprint arXiv:2606.11109, on PithPith's reviewarXiv
- [13] Taewook Nam, Junmo Cho, Youngsoo Jang, and Sung Ju Hwang. SpeedAug: Policy acceleration via tempo-enriched policy and RL fine-tuning.arXiv preprint arXiv:2512.00062, on PithPith's reviewarXiv
- [14] Octo Model Team, Dibya Ghosh, Homer Walke, Karl Pertsch, Kevin Black, Oier Mees, Sudeep Dasari, Joey Hejna, Tobias Kreiman, Charles Xu, et al. Octo: An open-source generalist robot policy.arXiv preprint arXiv:2405.12213, on PithPith's reviewarXiv
- [15] Jyothish Pari, Nur Muhammad Shafiullah, Sridhar Pandian Arunachalam, and Lerrel Pinto. The surprising effectiveness of representation learning for visual imitation.arXiv preprint arXiv:2112.01511, on PithPith's reviewarXiv
- [19] Chen Yang, Yucheng Hu, Yunchao Ma, Yunhuan Yang, Jing Tan, and Haoqiang Fan. Realtime-VLA V2: Learning to run VLAs fast, smooth, and accurate.arXiv preprint arXiv:2603.26360, on PitharXiv
- [20] 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, on PithPith's reviewarXiv
- [22] Yuke Zhu, Josiah Wong, Ajay Mandlekar, Roberto Martín-Martín, Abhishek Joshi, Kevin Lin, Abhiram Maddukuri, Soroush Nasiriany, and Yifeng Zhu. robosuite: A modular simulation framework and benchmark for robot learning. arXiv preprint arXiv:2009.12293, on PithPith's reviewarXiv
- [23] 14 Appendix A Hardware Setup The real-robot experiments are conducted on a Rokae Helios series humanoid robot, using one of its two seven-degree-of-freedom arms equipped with a Daimon gripper. The arm operates in Cartesian impedance mode and executes policy-predicted end-effec... on Pith
- [24] per task using the task’s original demonstrations, without speed-related processing. Forπ0.5 and Fast-WAM, we use the released checkpoints without additional training or fine-tuning (Physical Intelligence, 2025; Yuan et al., 2026). All policy weights remain frozen throughout e... on Pith
- [27] using end-effector pose increments (Aljalbout et al., 2024). The controller clips each action dimension independently to[−1, 1]and linearly maps it to translation increments of up to0.05m or rotation increments of up to0.5rad per component. Such element-wise clipping can alter... on Pith
- [28] pretrained through behavior cloning and released by the DPPO authors (Ren et al., 2025). We use these checkpoints without RL fine-tuning and keep their weights frozen throughout evaluation. We evaluate Lift, Can, and Square using20denoising steps. The reference achieves74.0% s... on Pith
- [1987] Byungju Kim, Jinu Pahk, Chungwoo Lee, Jaejoon Kim, Jangha Lee, Theo Taeyeong Kim, Kyuhwan Shim, Jun Ki Lee, and Byoung-Tak Zhang. ESPADA: Execution speedup via semantics aware demonstration data downsampling for imitation learning.arXiv preprint arXiv:2512.07371, 2025a. Moo Ji... on PithPith's reviewarXiv
- [2011] Shilin Shan, Chuhao Zhou, Ruize Wang, Xinyan Chen, Xiangyu Chen, Xinyu Zhou, Boyu Ma, Iris Yuxuan Hu, Jingliang Li, Celeste Yuxuan Hu, et al. Learning physical interaction: A survey of tactile-and force-aware robot learning.arXiv preprint arXiv:2608.07558, on PithPith's reviewarXiv
- [2012] Zhichao Wu, Junyin Ye, Zhilong Zhang, Yihao Sun, Haoxin Lin, Jiaheng Luo, Haoxiang Ren, Lei Yuan, and Yang Yu. Speedup patch: Learning a plug-and-play policy to accelerate embodied manipulation.arXiv preprint arXiv:2603.20658, on PitharXiv
- [2016] with language-conditioned FiLM (Perez et al., 2018; Brohan et al., 2022; Jang et al., 2022; Haldar et al., 2024; Kim et al., 2025b) and spatial-softmax pooling (Levine et al., 2016). Each of the six modalities is encoded into 64 dimensions, and the features are concatenated ac... on Pith
show 9 more
- [2017] to10−5. Inference uses ten explicit Euler steps. For the real-robot checkpoint, generating one action chunk takes22.9± 0.45 ms on an NVIDIA RTX 5090 at batch size one using FP32. This timing includes observation encoding but excludes preprocessing, communication, and BERT comp... on Pith
- [2018] Physical Intelligence. π0.5: A vision-language-action model with open-world generalization.arXiv preprint arXiv:2504.16054, on PithPith's reviewarXiv
- [2019] Association for Computational Linguistics. 11 Junqiao Fan, Yuxuan Hu, Bofan Lyu, Yanshuo Lu, Pengfei Liu, Jiarui Zhang, Fangqiang Ding, Lihua Xie, Gen Li, and Jianfei Yang. mmHRI: Towards privacy-preserving human-robot interaction with millimeter-wave radar.arXiv preprint arXi... on PithPith's reviewarXiv
- [2020] Jiyao Zhang, Zimu Han, Junhan Wang, Xionghao Wu, Shihong Lin, Jinzhou Li, Hongwei Fan, Ruihai Wu, Dongjiang Li, and Hao Dong. HiPolicy: Hierarchical multi-frequency action chunking for policy learning.arXiv preprint arXiv:2604.06067, on PithPith's reviewarXiv
- [2022] Kang Chen, Zhihao Liu, Tonghe Zhang, Zhen Guo, Si Xu, Hao Lin, Hongzhi Zang, Xiang Li, Quanlu Zhang, Zhaofei Yu, et al.πRL: Online RL fine-tuning for flow-based vision-language-action models.arXiv preprint arXiv:2510.25889, on PitharXiv
- [2023] Kevin Black, Noah Brown, Danny Driess, Adnan Esmail, Michael Equi, Chelsea Finn, Niccolo Fusai, Lachy Groom, Karol Hausman, Brian Ichter, Szymon Jakubczak, Tim Jones, Liyiming Ke, Sergey Levine, Adrian Li-Bell, Mohith Mothukuri, Suraj Nair, Karl Pertsch, Lucy Xiaoyang Shi, Jam... on PithPith's reviewarXiv
- [2024] Anthony Brohan, Noah Brown, Justice Carbajal, Yevgen Chebotar, Joseph Dabis, Chelsea Finn, Keerthana Gopalakr- ishnan, Karol Hausman, Alex Herzog, Jasmine Hsu, et al. RT-1: Robotics transformer for real-world control at scale.arXiv preprint arXiv:2212.06817, on PithPith's reviewarXiv
- [2025] Jiaqi Bai, Jindou Jia, Yuxuan Hu, Gen Li, Xiangyu Chen, Tuo An, Kuangji Zuo, and Jianfei Yang. FLASH: Efficient visuomotor policy via sparse sampling.arXiv preprint arXiv:2605.15492, on PithPith's reviewarXiv
- [2026] Soumya Suvra Ghosal, Yiyou Sun, and Yixuan Li. How to overcome curse-of-dimensionality for out-of-distribution detection?arXiv preprint arXiv:2312.14452, on PithPith's reviewarXiv
Cited by 1 Pith paper
Reviewed papers on Pith that cite this one. The full citation record
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.
Come back by email
A short list of the week's newly reviewed papers. Nothing is sent until you confirm from your inbox, and every email has a link that stops it. No account.
No comments yet.
Sign in to comment