Pith.

Robotics · arXiv:2610.37378 v1 · submitted 27 Oct 2026 · 35 references

VLA-Stem: A Lightweight Safety Mechanism for Vision-Language-Action Models via Reinforcement Learning

Bojun Zou, Wenhao Yu, Jiaohao Li, Jiaxuan Lin, Guanting Ye, Haohan Geng, Weihao Yu, Yanyong Zhang

show 1 more author

Jianmin Ji

The paper says

“Vision-Language-Action (VLA) models integrate high-level reasoning with low-level action execution for end-to-end robotic control. However, most VLA models target static manipulation in unconstrained environments, overlooking spatial constraints and dynamic obstacles.”

more of the abstract

Fine-tuning VLAs on obstacle-avoidance data is a straightforward solution, but collecting such data is costly, especially for dynamic scenarios, and may degrade their original task capabilities. To address these limitations, we propose VLA-Stem, a lightweight, reinforcement-learning-based safety mechanism for VLA systems. Without fine-tuning the underlying VLA policy, VLA-Stem converts its action chunks into motion references and uses trajectory-level guidance to preserve their spatial intent while allowing flexible motion and timing adjustments. Conditioned on these references, robot proprioception, and local obstacle information, the Stem policy adaptively modifies execution at inference time, preserving the original task intent while enabling responsive adaptation to spatial constraints and dynamic obstacles. For robust and sample-efficient training, we combine teacher-student distillation with curriculum learning. We evaluate VLA-Stem on reference following, static spatial constraints, and dynamic obstacle avoidance. Across six constrained simulation settings, VLA-Stem achieves 52.8% average collision-free task success versus 23.7% for the strongest safety baseline, while preserving base-policy performance in obstacle-free settings; real-world experiments further show consistent collision reductions under static and dynamic obstacles. These results demonstrate that VLA-Stem couples high-level VLA task intent with rapid, reflex-like responses to immediate physical risks, providing a brainstem-inspired mechanism for robust robot execution.

from the abstract

Pith says

Machine review · deepseek-v4-flash · 28 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. A clean, well-ablated safety wrapper for frozen VLAs; the trajectory-level guidance is the real novelty, but the headline margin needs seeds and error bars before it is established. This line is wrong

4 major objections · 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 35 references say.

The paper says, at §IV-B, Table I with §IV-A, Table VI

The review did not quote this passage.

Pith says: Objection 1 of 4 · major

Every entry in Table I is a single point estimate. The training-budget table reports transitions 'per seed' and the text describes 'the evaluated checkpoints' in the singular form, which, combined with no stated seed count, indicates one PPO training run per policy and scenario.

more of objection 1

Because each setting is evaluated over 3,200 episodes, the within-policy standard error is roughly 0.01, so the missing variation is the between-seed variance of PPO training. The headline margin (52.8% versus 23.7%) can therefore not be assessed: it may be genuine, or it may shrink substantially under another seed. Please report results over at least three to five seeds (mean ± std or interquartile range) for the main comparison, state the number of seeds for every reported entry, and, if only single-seed runs were performed, say so explicitly and soften the headline claim accordingly; releasing the trained checkpoints would make the margin independently verifiable.

The paper says, at §III-A, Eq. (2)

The review did not quote this passage.

Pith says: Objection 2 of 4 · major

The conversion of the VLA action chunk into future end-effector poses is underspecified. The paper does not state whether {T^w_{t_c+k}} is obtained by forward kinematics applied to raw joint targets, by a simulator or dynamics roll-out of the chunk, or by an assumed perfect-tracking model, nor how the conversion is implemented on the real robot, where the base VLA and the Stem act in different action spaces (joint-based VLA output versus end-effector delta-pose Stem output).

more of objection 2

This conversion is the only channel carrying task intent into the reference τ_rel^c, so its fidelity is load-bearing for the whole method. Please specify the procedure and report a quantitative check of its accuracy, for example the open-loop endpoint error of the reference relative to the executed trajectory.

The paper says, at §IV-C, Fig. 5

The review did not quote this passage.

Pith says: Objection 3 of 4 · major

The real-world evaluation reports only aggregate rates over 15 trials per task and scenario. With 15 binomial trials, 95% confidence intervals are on the order of ±20–25 percentage points around the reported success rates, so the displayed differences (0–20% to 47–60%) are much less precise than the figure suggests, and the adjudication of success and collision is not described (human labeling, thresholds on AprilTag or vision estimates, or both).

more of objection 3

Please report raw per-trial counts and the adjudication protocol per metric; the real-world results are one of the two headline supports, and this information is needed to interpret them.

The paper says, at Abstract and §IV-B, Table I

The review did not quote this passage.

Pith says: Objection 4 of 4 · major

The claim '23.7% for the strongest safety baseline' is the per-setting best among the two VLSA variants only. Averaging the per-setting best over all baselines (VLSA-EE, VLSA-ARM, VLA-finetune) across the same six constrained settings from Table I yields 32.4%, and VLA-finetune is the strongest baseline in four of the six settings.

more of objection 4

The conclusion that VLA-Stem has the highest SUCC is unaffected, but the headline figure as stated understates the strongest competition; either report the comparison against the per-setting best of all baselines or define explicitly that the 23.7% refers to the plug-and-play VLSA family.

5 minor objections

The paper says, at App. C, Eq. (23)

Pith says: minor objection 1 of 5

α is printed as 'α = 47', but the weighted form d̃_i = 0.7 d_i = 0.7 d_pos_i + 0.4 d_rot_i directly below it is consistent only with α = 4/7; as printed, the guidance reward is not reproducible. Please correct the typo and consider reporting sensitivity of the main results to α.

The paper says, at §IV-B2, Fig. 4

Pith says: minor objection 2 of 5

The perturbation experiment largely confirms that the policy optimizes the guidance reward of Eqs. (7) and (27); the sentence claiming that it 'demonstrates that responsive Stem execution remains coupled with subsequent VLA task progression' is stronger than the 30-rollout-per-direction evidence can establish. Frame the experiment as a sanity check of the conditioning mechanism and rest the coupling claim on Tables III–IV.

show 3 more

The paper says, at §IV-A

Pith says: minor objection 3 of 5

No code, data, or checkpoint release is indicated. Given that the central numbers are single-seed point estimates, a release (or seed-averaged statistics) is needed for independent verification of the headline margin.

The paper says, at §IV-A, Table I

Pith says: minor objection 4 of 5

The relationship between SUCC and COLL should be stated explicitly. Since the two columns do not sum to 1 in Table I (e.g., PickCube-Static for VLA-Stem: 0.5219 + 0.2316 = 0.7535), an episode can apparently collide and still complete the task, and episodes can fail without collision; defining whether COLL counts all violating episodes or only those that also fail, and whether SUCC requires zero violations by definition, would remove ambiguity.

The paper says, at §IV-A

Pith says: minor objection 5 of 5

The discount factor γ = 0.8 with a 100-step rollout horizon heavily discounts rewards beyond roughly 10–15 steps; because dynamic-scenario success explicitly requires waiting for moving obstacles, a sentence justifying this choice or an ablation on γ would strengthen the training-details section.

What would settle it

Release code and rerun Table I for VLA-Stem, VLSA-EE, VLSA-ARM, and VLA-finetune from three independently seeded training runs, using the same budgets and the same 3,200-episode evaluation per setting. The concern is resolved if, for every seed, VLA-Stem's six-setting average SUCC exceeds every baseline's average SUCC and VLA-Stem is best in every constrained setting. It is not resolved if the margins overlap across seeds or any baseline wins in any seed.

from the skeptic's reading

How the review went

4 readings by deepseek-v4-flash · 28 Oct 2026

  1. Reader holds up, with conditions confidence moderate
  2. Skeptic partly agrees 3 min 33 s · 18,878 tokens read, 24,343 written
  3. Editor wrote the verdict above 55 s · 20,400 tokens read, 863 written
  4. Referee major revision: 4 major objections, 5 minor 2 min 5 s · 19,224 tokens read, 16,615 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

I basically agree with the reader. The core idea is clean: freeze the VLA, convert each action chunk into an end-effector trajectory reference, and train a small RL policy that conditions on that reference plus proprioception and obstacle geometry. The genuinely new piece is the trajectory-level guidance objective—matching the current pose to a forward corridor of reference segments rather than penalizing per-step action deviation—and they isolate it with the right ablation: replacing it with step-wise action alignment drops average success from 63% to 30%. The other ablations do their job too: removing the reference collapses success, and removing distillation hurts. The normal-scenario numbers show the wrapper does not obviously degrade the base policy. That is a real, practically useful result, and the experimental design is mostly on the level: 3,200 episodes in simulation, the same base checkpoint for all safety wrappers, and an honest discussion of why low collision rate does not imply task success.

The soft spot is the one the stress test flags. The headline claim is a comparison of point estimates—52.8% versus 23.7%—and I cannot tell how stable it is. Table VI says transitions "per seed," there is no seed count, no standard deviations, no significance tests, and no code or data release. PPO runs vary; a 29-point margin might survive another seed or might shrink. This does not undercut the internal coherence of the method, and the ablations are consistent with the mechanism, but it does mean the strongest empirical assertion is underdetermined. I would not call this fatal; I would call it the thing to fix before the number is quoted as established.

Minor things: VLA-finetune is reported at its best demonstration budget, which is defensible but should be stated with its validation selection; the real-world evaluation is 15 trials per setting, small but directionally consistent; and obstacle annotations come from AprilTags rather than learned perception, which the authors themselves scope in the conclusion. The end-effector reference may not transfer to contact- or force-sensitive tasks—also acknowledged.

Who gets value: people working on test-time safety for VLA policies. I would send it to review, with the request that the authors add multi-seed runs, error bars, and ideally code or data. The mechanism is worth refereeing; the numbers just need to be anchored.

The strongest objection

The paper's headline is a comparative numerical claim: VLA-Stem averages 52.8% collision-free success across six constrained settings versus 23.7% for the strongest baseline. Table I reports one point estimate per condition with no seed count, standard deviation, confidence interval, or significance test, and no code or data release is indicated. The training budget table (Table VI) describes transitions 'per seed,' suggesting each reported policy may come from a single training run rather than a distribution over seeds. Because PPO training is stochastic and the final policy can shift by several points across seeds, the reported 29-point margin cannot be assessed without knowing its variability: the method could be genuinely superior, or the margin could shrink under another seed. This is not an internal inconsistency; the ablations for reference removal, trajectory-level guidance, and teacher-student distillation are internally coherent and support the proposed mechanism. The concern is that the strongest empirical assertion, as stated, is underdetermined by the presented evidence.

The reader's CONDITIONAL verdict is appropriate. The method itself is coherent: the reference-conditioned formulation, trajectory-level guidance, and distillation ablations all support the proposed mechanism, and the paper's own perturbed-reference experiment partially addresses concern about reference fidelity. My most load-bearing concern is different from the reader's named weakest assumption: rather than the sufficiency of pose references, the unresolved issue is the stability of the headline quantitative comparison. Table I gives no seed count, no confidence intervals, and no released artifacts, so the 52.8% versus 23.7% margin is a point estimate whose variability is unknown. The reader did note the absence of error bars and code in the rationale, which is why agreement is partial rather than full. I recommend keeping the verdict CONDITIONAL: if the three-seed check confirms the margins, the quantitative claim would be substantially strengthened; if not, the headline comparison would need to be weakened or the paper moved toward unverified.

The referee's report

VLA-Stem addresses a real gap—runtime safety for frozen VLAs without fine-tuning or candidate search—and the design is well motivated: the trajectory-level guidance and forward-corridor matching are cleanly formulated (Sec. III-B, App. C), and the ablation set is internally coherent and supports the claimed mechanism: removing the reference collapses success (Table II), stepwise alignment degrades the safety-task trade-off (Table III), and removing distillation drops success and raises collisions (Table IV). The trajectory-intervention probe of Fig. 4 is a thoughtful direct test of reference conditioning, and the sim-to-real transfer without real-world training is a genuine practical strength. The authors explicitly acknowledge the scope limits (structured tasks, low-dimensional obstacle inputs, no contact/force objectives, reactive rather than predictive baselines). The principal weakness is statistical: the headline comparison rests on single-seed point estimates with no variance characterization, and the current evidence cannot rule out a substantially smaller margin under training-seed variation.

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

The paper claims that a vision-language-action (VLA) robot policy, which turns camera images plus a language instruction into end-effector actions, can be made collision-safe at inference time without retraining the base model. VLA-Stem adds a small reinforcement-trained 'Stem' policy that converts each VLA action chunk into a short reference trajectory and then executes adapted versions of it, rewarded for staying spatially near the reference while freely changing the local path and its timing. In six constrained simulation settings the authors report 52.8% average collision-free task success versus 23.7% for the strongest compared safety layer, with obstacle-free success matching or slightly exceeding the base policy; real-world stowing and stacking trials cut collision rates from 73–100% to 20–40%. A reader should care because obstacle avoidance is normally bought with expensive avoidance data or with safety filters that disrupt the task, whereas VLA-Stem separates task intent (frozen VLA) from reflexes (Stem) and trains only the small reflex layer.

What carries the argument. The reference trajectory is the carrier of task intent: at the start of each 8-step interval the frozen VLA emits an action chunk, which is converted into future end-effector poses expressed relative to the current pose (Eq. (2)), and the Stem policy conditions on this reference plus proprioception and local obstacle bounding boxes (Eqs. (3)–(4)). The load-bearing objective is the trajectory-level guidance term (Eqs. (5)–(7)), which projects the current end-effector pose onto reference segments, restricts matching to a non-decreasing forward corridor with look-ahead $K$, and rewards closeness in a combined Euclidean-position and quaternion-geodesic-rotation measure. This is what allows re-timing and local detours without punishment for lost step-wise alignment, and it is paired with teacher–student distillation (Eq. (8)) and a three-stage curriculum to make recovery from avoidance-induced deviations learnable. The whole Stem policy is a small MLP actor–critic of roughly 4.56 million parameters trained with PPO, deployed with only low-dimensional inputs and no raw vision.

Core claim. The central claim is that a VLA policy's action chunk, converted into a relative end-effector pose trajectory (Eq. (2)), is a faithful and cheap carrier of task intent, and that a separate policy rewarded to follow that trajectory through a forward-moving corridor (Eqs. (5)–(7)) can preserve the task while dodging obstacles. The paper calls this trajectory-level guidance: instead of matching the VLA action at each time step, the Stem policy is scored by how far the executed end-effector pose falls from the nearby segments of the reference, with segment matching restricted so the progress index never moves backward, which lets the robot wait or detour without accumulating time-alignment penalties. Training is staged from reference following to static constraints to moving obstacles, and a teacher that sees privileged task state such as target-object positions is distilled into a deployable student that sees only proprioception, obstacle boxes, and the reference. The ablations make the load-bearing role of each piece explicit: removing the reference collapses task success to zero, replacing trajectory-level guidance with step-wise action matching cuts average success from 63% to 30%, and removing distillation drops average constrained success from 63% to 14% while raising collision rates.

The premise it rests on. The load-bearing premise is that the short end-effector trajectory implied by the VLA action chunk carries enough of the task's intent that staying spatially near it while detouring and waiting still completes the task, together with the requirement of accurate local obstacle positions at every step; the authors acknowledge the trajectory-as-intent assumption is untested for contact-rich or force-sensitive manipulation and for noisy perception.

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

Obstacle safety can be added to a frozen VLA purely at inference time: the base checkpoint is never updated, so a new task needs a Stem training run, not VLA fine-tuning data.

Under static and dynamic obstacles, collision-free task success more than doubles on average relative to the strongest compared safety layer (52.8% versus 23.7% across the six constrained settings).

The specific mechanism matters: replacing trajectory-level guidance with step-wise action alignment halves average success (63% to 30%), so the forward-corridor objective, not the task reward, is what preserves behaviour during detours (Table III).

Recovery from avoidance-induced deviations depends on privileged task information during training: distillation raises average constrained success from 14% to 63% and lowers the average collision rate from 37% to 15% (Table IV).

In obstacle-free scenes the safety layer does not degrade the base policy: observed success matches or slightly exceeds the base $\pi_{0.5}$ on all three tasks (0.7325 vs 0.7166; 0.7061 vs 0.6138; 0.7521 vs 0.7450).

Because the only interface to the frozen model is its action chunk, the same Stem recipe should attach to any action-chunk policy, including diffusion policies and other embodied foundation models, provided their outputs convert to end-effector poses; the paper demonstrates this only on $\pi_0$ and $\pi_{0.5}$.

The 8-step reference horizon with look-ahead $K = 6$ sets a fixed ceiling on how long the policy can wait or detour before the forward corridor is exhausted, so obstacle speeds that force holds beyond one chunk would presumably require re-querying the VLA or longer chunks — a threshold the paper does not measure.

The deployable student takes deliberately low-dimensional inputs with no raw vision, which makes high-rate control plausible; the practical bottleneck is perception, so replacing the fiducial-marker obstacle input with learned detection is a direct test of whether the sim-to-real gap stays small.

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

Vision-Language-Action (VLA) models integrate high-level reasoning with low-level action execution for end-to-end robotic control. However, most VLA models target static manipulation in unconstrained environments, overlooking spatial constraints and dynamic obstacles. Fine-tuning VLAs on obstacle-avoidance data is a straightforward solution, but collecting such data is costly, especially for dynamic scenarios, and may degrade their original task capabilities. To address these limitations, we propose VLA-Stem, a lightweight, reinforcement-learning-based safety mechanism for VLA systems. Without fine-tuning the underlying VLA policy, VLA-Stem converts its action chunks into motion references and uses trajectory-level guidance to preserve their spatial intent while allowing flexible motion and timing adjustments. Conditioned on these references, robot proprioception, and local obstacle information, the Stem policy adaptively modifies execution at inference time, preserving the original task intent while enabling responsive adaptation to spatial constraints and dynamic obstacles. For robust and sample-efficient training, we combine teacher-student distillation with curriculum learning. We evaluate VLA-Stem on reference following, static spatial constraints, and dynamic obstacle avoidance. Across six constrained simulation settings, VLA-Stem achieves 52.8% average collision-free task success versus 23.7% for the strongest safety baseline, while preserving base-policy performance in obstacle-free settings; real-world experiments further show consistent collision reductions under static and dynamic obstacles. These results demonstrate that VLA-Stem couples high-level VLA task intent with rapid, reflex-like responses to immediate physical risks, providing a brainstem-inspired mechanism for robust robot execution.

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 · 2026-10-28

The authors' main claim, as the classifier picked it out 0.96

“These results demonstrate that VLA-Stem couples high-level VLA task intent with rapid, reflex-like responses to immediate physical risks, providing a brainstem-inspired mechanism for robust robot execution.”

How the main claim is stated: asserted 1.00

What the abstract itself states, as probabilities: claims first 0.05 · states limitation 0.16 · human participants 0.03 · controlled comparison 0.96 · randomized assignment 0.02 · states quantitative result 0.99 · compares with prior results 0.97 · 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.

SentenceRoleThe authors' own workStates the main conclusionStrength
Vision-Language-Action (VLA) models integrate high-level reasoning with low-level action execution for end-to-end robotic control. background 1.00 0.05 0.02 asserted 1.00
However, most VLA models target static manipulation in unconstrained environments, overlooking spatial constraints and dynamic obstacles. gap 0.92 0.07 0.02 asserted 1.00
Fine-tuning VLAs on obstacle-avoidance data is a straightforward solution, but collecting such data is costly, especially for dynamic scenarios, and may degrade their original task capabilities. background 0.56 0.09 0.02 asserted 0.65
To address these limitations, we propose VLA-Stem, a lightweight, reinforcement-learning-based safety mechanism for VLA systems. contribution 0.99 0.73 0.06 asserted 0.97
Without fine-tuning the underlying VLA policy, VLA-Stem converts its action chunks into motion references and uses trajectory-level guidance to preserve their spatial intent while allowing flexible motion and timing adjustments. method 0.98 0.77 0.06 asserted 1.00
Conditioned on these references, robot proprioception, and local obstacle information, the Stem policy adaptively modifies execution at inference time, preserving the original task intent while enabling responsive adaptation to spatial constraints and dynamic obstacles. method 0.99 0.63 0.11 asserted 1.00
For robust and sample-efficient training, we combine teacher-student distillation with curriculum learning. method 0.99 0.89 0.03 asserted 0.98
We evaluate VLA-Stem on reference following, static spatial constraints, and dynamic obstacle avoidance. method 0.71 0.94 0.03 asserted 0.88
Across six constrained simulation settings, VLA-Stem achieves 52.8% average collision-free task success versus 23.7% for the strongest safety baseline, while preserving base-policy performance in obstacle-free settings; real-world experiments further show consistent collision reductions under static and dynamic obstacles. result 1.00 0.97 0.32 asserted 1.00
These results demonstrate that VLA-Stem couples high-level VLA task intent with rapid, reflex-like responses to immediate physical risks, providing a brainstem-inspired mechanism for robust robot execution. interpretation 1.00 0.72 0.88 asserted 1.00

What kind of paper the classifier reports this is: method or tool 1.00 · evidence from observation or experiment 0.84

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, 2026-10-28: 1 request, 7,678 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 (5)
Figure 1 of the paper
Figure 1. This figure illustrates the biological motivation behind our method.
Figure 2 of the paper
Figure 2. VLA-Stem pipeline. The teacher is trained using curriculum learning, progressing from reference trajectory following to static obstacle avoidance and then to dynamic obstacle avoidance, with gradually increasing difficulty. The teacher policy takes as input the trajectory sequence produced by the base VLA policy, robot proprioception, and obstacle information, and additionally leverages privileged task information about the target object to accelerate training. Finally, the teacher policy is used to distill a student policy, removing the dependence on task-specific privileged…
Figure 4 of the paper
Figure 4. Reference-conditioned execution on StackCube: trajectory
Figure 5 of the paper
Figure 5. Real-world evaluation of VLA-Stem. (a) Real-world experimental examples. (b–c) Quantitative results on Object Stowing and Stacking, respectively.

Showing 4 of the paper's 5 figures. Figures are the authors', reproduced from the paper.

References (35)
  1. [1] Physical Intelligence, K. Black, N. Brown, J. Darpinian, K. Dhabalia, D. Driess, A. Esmail, M. Equi, C. Finn, N. Fusai,et al., “ π0.5: a vision-language-action model with open-world generalization,”arXiv preprint arXiv:2504.16054, 2025 on PithPith's reviewarXiv
  2. [2] J. Yang, J. J. Liu, Y . Li, Y . Khaky, K. Shaw, and D. Pathak, “Deep reactive policy: Learning reactive manipulator motion planning for dynamic environments,”arXiv preprint arXiv:2509.06953, 2025 on PithPith's reviewarXiv
  3. [3] S. Hu, Z. Liu, S. Liu, J. Cen, Z. Meng, S. Wang, X. Li, and X. He, “VLSA: Vision-Language-Action models with plug-and-play safety constraint layer,”arXiv preprint arXiv:2512.11891, 2025 on PithPith's reviewarXiv
  4. [4] G. Ye, Q. Zhao, W. Yu, X. Zhang, J. Ji, Y . Zhang, and K.-V . Yuen, “C2RoPE: Causal continuous rotary positional encoding for 3d large multimodal-models reasoning,”arXiv preprint arXiv:2602.10551, 2026 on PitharXiv
  5. [5] K. Yip, Q. Zhao, W. Yu, L. Yuen, M. Li, X. Zhang, J. Ji, Y . Zhang, Q. Jiang, and K.-V . Yuen, “Sope: Spherical coordinate-based positional embedding for enhancing spatial perception of 3d lvlms,” inProceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recogn... on Pith
  6. [6] M. J. Kim, K. Pertsch, S. Karamcheti, T. Xiao, A. Balakrishna, S. Nair, R. Rafailov, E. Foster, G. Lam, P. Sanketi,et al., “OpenVLA: An open- source vision-language-action model,”arXiv preprint arXiv:2406.09246, 2024 on PithPith's reviewarXiv
  7. [7] B. Zitkovich, T. Yu, S. Xu, P. Xu, T. Xiao, F. Xia, J. Wu, P. Wohlhart, S. Welker, A. Wahid,et al., “RT-2: Vision-language-action models transfer web knowledge to robotic control,” inConference on Robot Learning. PMLR, 2023, pp. 2165–2183 on Pith
  8. [8] A. Brohan, N. Brown, J. Carbajal, Y . Chebotar, J. Dabis, C. Finn, K. Gopalakrishnan, K. Hausman, A. Herzog, J. Hsu,et al., “RT-1: Robotics transformer for real-world control at scale,”arXiv preprint arXiv:2212.06817, 2022 on PithPith's reviewarXiv
  9. [9] S. Liu, L. Wu, B. Li, H. Tan, H. Chen, Z. Wang, K. Xu, H. Su, and J. Zhu, “RDT-1B: a diffusion foundation model for bimanual manipulation,”arXiv preprint arXiv:2410.07864, 2024 on PithPith's reviewarXiv
  10. [10] K. Black, N. Brown, D. Driess, A. Esmail, M. Equi, C. Finn, N. Fusai, L. Groom, K. Hausman, B. Ichter,et al., “ π0: A vision-language-action flow model for general robot control,”arXiv preprint arXiv:2410.24164, 2024 on PithPith's reviewarXiv
  11. [11] Y . Xiong, W. Yu, J. Lin, B. Zou, J. Li, L. Zhang, Y . Zhang, and J. Ji, “Cac-vla: Context-gated action conditioning for vision-language-action models,”arXiv preprint arXiv:2607.04816, 2026 on PithPith's reviewarXiv
  12. [12] H. Geng, S. Zhang, W. Sun, G. Ye, W. Yu, Y . Xie, Q. Zeng, P. Ye, J. Deng, and J. Ji, “Trajwam: An efficient trajectory-guided world action model for dynamic robotic manipulation,”arXiv preprint arXiv:2610.25763, 2026 on PithPith's reviewarXiv
  13. [13] G. Lu, C. Zhang, H. Jiang, Y . Zhou, Z. Gao, Y . Tang, and Z. Wang, “VLA-RL: Towards masterful and general robotic manipulation with scalable reinforcement learning,” 2025, Notion Blog on Pith
  14. [14] A. Amin, R. Aniceto, A. Balakrishna, K. Black, K. Conley, G. Connors, J. Darpinian, K. Dhabalia, J. DiCarlo, D. Driess,et al., “ π0.6: a VLA that learns from experience,”arXiv preprint arXiv:2511.14759, 2025 on PithPith's reviewarXiv
  15. [15] S. M. LaValle and J. J. Kuffner, “Rapidly-exploring random trees: Progress and prospects,”Algorithmic and Computational Robotics, pp. 303–307, 2001 on Pith
  16. [16] L. E. Kavraki, P. Svestka, J.-C. Latombe, and M. H. Overmars, “Prob- abilistic roadmaps for path planning in high-dimensional configuration spaces,”IEEE Transactions on Robotics and Automation, vol. 12, no. 4, pp. 566–580, 1996 on Pith
  17. [17] N. Ratliff, M. Zucker, J. A. Bagnell, and S. Srinivasa, “CHOMP: Gradient optimization techniques for efficient motion planning,” in 2009 IEEE International Conference on Robotics and Automation. IEEE, 2009, pp. 489–494 on Pith
  18. [18] J. Schulman, Y . Duan, J. Ho, A. Lee, I. Awwal, H. Bradlow, J. Pan, S. Patil, K. Goldberg, and P. Abbeel, “Motion planning with sequential convex optimization and convex collision checking,”The International Journal of Robotics Research, vol. 33, no. 9, pp. 1251–1270, 2014 on Pith
  19. [19] M. Kalakrishnan, S. Chitta, E. Theodorou, P. Pastor, and S. Schaal, “STOMP: Stochastic trajectory optimization for motion planning,” in 2011 IEEE International Conference on Robotics and Automation. IEEE, 2011, pp. 4569–4574 on Pith
  20. [20] W. Yu, J. Peng, H. Yang, J. Zhang, Y . Duan, J. Ji, and Y . Zhang, “Ldp: A local diffusion planner for efficient robot navigation and collision avoidance,” in2024 IEEE/RSJ international conference on intelligent robots and systems (IROS). IEEE, 2024, pp. 5466–5472 on Pith
show 15 more
  1. [21] J. Carvalho, A. T. Le, M. Baierl, D. Koert, and J. Peters, “Motion planning diffusion: Learning and planning of robot motions with diffusion models,” in2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2023, pp. 1916–1923 on Pith
  2. [22] F. Ying, H. Zhang, H. Wang, H. Huang, and M. H. Ang Jr, “URPlanner: A universal paradigm for collision-free robotic motion planning based on deep reinforcement learning,”arXiv preprint arXiv:2505.20175, 2025 on PithPith's reviewarXiv
  3. [23] W. Yu, J. Peng, Q. Qiu, H. Wang, L. Zhang, and J. Ji, “Pathrl: An end-to-end path generation method for collision avoidance via deep reinforcement learning,” in2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2024, pp. 9278–9284 on Pith
  4. [24] R. Cheng, G. Orosz, R. M. Murray, and J. W. Burdick, “End-to-end safe reinforcement learning through barrier functions for safety-critical continuous control tasks,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 33, no. 01, 2019, pp. 3387–3395 on Pith
  5. [25] J. Breeden and D. Panagou, “Predictive control barrier functions for online safety critical control,” in2022 IEEE 61st Conference on Decision and Control (CDC). IEEE, 2022, pp. 924–931 on Pith
  6. [26] B. Zhang, Y . Zhang, J. Ji, Y . Lei, J. Dai, Y . Chen, and Y . Yang, “SafeVLA: Towards safety alignment of vision-language-action model via constrained learning,”Advances in Neural Information Processing Systems, vol. 38, pp. 153 335–153 373, 2026 on Pith
  7. [27] W. Jung, D. Anthony, U. A. Mishra, N. R. Arachchige, M. Bronars, D. Xu, and S. Kousik, “RAIL: Reachability-aided imitation learning for safe policy execution,” in2025 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2025, pp. 3582–3589 on Pith
  8. [28] H. Deng, Z. Wu, H. Liu, W. Guo, Y . Xue, Z. Shan, C. Zhang, B. Jia, Y . Ling, G. Lu,et al., “A survey on reinforcement learning of Vision- Language-Action models for robotic manipulation,”Authorea Preprints, 2025 on Pith
  9. [29] W. Guo, G. Lu, H. Deng, Z. Wu, Y . Tang, and Z. Wang, “VLA- Reasoner: Empowering vision-language-action models with reasoning via online monte carlo tree search,”arXiv preprint arXiv:2509.22643, 2025 on PitharXiv
  10. [30] M. Nakamoto, O. Mees, A. Kumar, and S. Levine, “Steering your generalists: Improving robotic foundation models via value guidance,” arXiv preprint arXiv:2410.13816, 2024 on PithPith's reviewarXiv
  11. [31] Y . Ze, Z. Chen, J. P. Ara ´ujo, Z.-a. Cao, X. B. Peng, J. Wu, and C. K. Liu, “Twist: Teleoperated whole-body imitation system,”arXiv preprint arXiv:2505.02833, 2025 on PithPith's reviewarXiv
  12. [32] J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov, “Prox- imal policy optimization algorithms,”arXiv preprint arXiv:1707.06347, 2017 on PithPith's reviewarXiv
  13. [33] Y . Bengio, J. Louradour, R. Collobert, and J. Weston, “Curriculum learning,” inProceedings of the 26th Annual International Conference on Machine Learning, 2009, pp. 41–48 on Pith
  14. [34] S. Tao, F. Xiang, A. Shukla, Y . Qin, X. Hinrichsen, X. Yuan, C. Bao, X. Lin, Y . Liu, T.-k. Chan, Y . Gao, X. Li, T. Mu, N. Xiao, A. Gurha, V . N. Rajesh, Y . W. Choi, Y .-R. Chen, Z. Huang, R. Calandra, R. Chen, S. Luo, and H. Su, “ManiSkill3: GPU parallelized robotics simul... on Pith
  15. [35] M. Krogius, A. Haggenmiller, and E. Olson, “Flexible layouts for fiducial tags,” inProceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), October 2019. APPENDIX A. Implementation and Execution Details on Pith

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