Pith. sign in

REVIEW 4 major objections 5 minor 38 references

Trajectory Divergence Horizon Decision for Reliable Dual-Arm Surgical Subtask Manipulation

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read A divergence gate lifts surgical VLA success to 80% on tissue tasks

desk verdict Plausible test-time safety gate for surgical VLA, with a real benchmark and real gains in contact stages, but the evidence is statistically thin and the divergence proxy is never validated. read the letter →

arxiv 2608.09125 v1 pith:UZB52A5S submitted 2026-08-10 cs.RO

Mingwu Su , Guankun Wang , Jinsong Lin , Rulin Zhou , Ziyi Hao , Zhiwei Fang , Huxin Gao , Jiewen Lai
show 3 more authors
Jiazheng Wang Fan Zhang Hongliang Ren
This is my paper · ORCID
classification cs.RO
keywords surgicalroboticsVision-Language-Actionmodelsflowmatchingadaptiveexecutionhorizontrajectorydivergencedual-armmanipulationtest-timesafetygateopen-loop
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Surgical Vision-Language-Action (VLA) policies typically commit to a fixed number of open-loop action steps, and under millimeter-precision contact with deformable tissue those blind executions accumulate errors. This paper argues that the right remedy is not a better policy but an adaptive test-time gate: decide, after each action chunk is generated, how many of its steps can be trusted. The proposed Trajectory Divergence Horizon Decision (TDHD) generates a second trajectory from a slightly perturbed noise sample, measures the per-step divergence between the two, and truncates execution at the first unstable step, forcing replanning from fresh observations. On a real dual-arm surgical robot with 600 teleoperated demonstrations, the paper reports that TDHD raises end-to-end success from 55% to 60% on needle manipulation and from 55% to 80% on tissue manipulation, with the largest gains in the contact-heavy final stages. The claim is that a lightweight divergence-based truncation rule can make fixed-horizon VLA policies reliable enough for surgical subtasks without retraining.

What carries the argument

The central object is the per-step trajectory divergence $\delta_i = \|A^{(1)}_i - A^{(2)}_i\|_2$, computed between two action trajectories generated by the same flow-matching velocity field from two noise samples $z_1$ and $z_2 = z_1 + \epsilon\eta$. Because flow-matching generation is deterministic given the noise and conditioning, this divergence isolates the model's sensitivity to tiny input perturbations, and the paper treats it as a proxy for prediction instability and execution risk, motivated by the classical idea that nearby trajectories diverging rapidly indicate instability. The dual-threshold stopping rule combines an absolute threshold $\theta_{\mathrm{abs}}$ on $\delta_i$ with a ratio threshold $r_{\mathrm{thresh}}$ on $\delta_i/(\delta_{i-1}+\xi)$, taking the earliest step where either is violated, and then constrains the final horizon to $[s_{\min}, s_{\max}]$ before dispatching actions to the robot.

What would settle it

One decisive check is to record, for the same action chunks, the step at which the true end-effector tracking error first exceeds a tolerance and compare it with TDHD's chosen truncation step; if truncation routinely happens after the error has already accumulated, the divergence proxy is not measuring execution risk. Another is to rerun the tissue trials with different phantom stiffness or target locations and see whether the reported tissue success rate holds; if the gain vanishes when the deformable material changes, the divergence signal is tuned to this benchmark's specific failure modes.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the reliability of a flow-matching VLA action chunk can be assessed at test time by comparing it with a twin trajectory generated from a slightly perturbed initial noise vector. The step-wise $\ell^2$-distance $\delta_i$ between the two trajectories serves as an instability score: small divergence means the action at step $i$ is safe to execute open-loop, large or sharply rising divergence means the model's prediction is unreliable and the robot should stop and replan. A dual-threshold rule selects the earliest unsafe step, and the executed horizon is clipped to a physical range $[s_{\min}, s_{\max}]$. On real hardware with 20 trials per setting, the paper shows this adaptive horizon control consistently beats fixed-horizon execution, with the clearest gains in needle pick/regrasp and tissue resection, where contact transitions cause the largest prediction drift.

Load-bearing premise

The load-bearing premise is that the divergence between two nearly identical generated action predictions tracks how likely those actions are to actually fail on the robot; if that proxy breaks down, the gate will truncate at the wrong times and the reported success-rate gains will not transfer beyond the paper's specific setup.

Editorial extensions

If this is right

  • If TDHD works as reported, fixed-horizon open-loop execution is identified as a major, addressable source of surgical VLA failure: the same policy with adaptive truncation outperforms any single fixed execution length in the ablation study.
  • The gains are stage-dependent: TDHD is most active during contact transitions such as needle pick and tissue resection, while stable phases like reach and lift execute close to the full horizon, so the gate behaves like a targeted safety intervention rather than a blanket slowdown.
  • Because TDHD only modifies how an already generated action chunk is executed, it can be applied on top of existing fine-tuned VLA policies to improve reliability without retraining or changing the policy weights.
  • The dual-threshold design means the gate responds both to sustained divergence and to sudden divergence jumps, allowing it to catch instability even when absolute divergence values remain modest.

Reading between the lines

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

  • The same divergence gate should transfer to any flow-matching or diffusion VLA policy whose action generation is deterministic given the initial noise, since the gate only needs two forward passes of the action decoder; the paper only reports results on one fine-tuned policy family, so this transfer is an extrapolation.
  • The divergence signal could also serve as an operator-facing risk monitor during semi-autonomous surgery: a rising $\delta_i$ marks the precise moments when a human should be ready to intervene, a use the paper does not discuss.
  • A natural extension the paper leaves implicit is per-stage threshold calibration; its own ablation shows task-dependent benefits, so tuning $\theta_{\mathrm{abs}}$ and $r_{\mathrm{thresh}}$ separately for reach, pick/lift, and regrasp/resection could yield further gains.
  • The gate's estimate uses only one perturbed trajectory; averaging divergence over several perturbations would reduce the variance of the instability estimate, and the paper's own future-work list mentions ensemble size as an open question.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes Trajectory Divergence Horizon Decision (TDHD), a test-time mechanism that adaptively decides how many steps of a predicted action chunk to execute before replanning. For each chunk, the policy generates two action trajectories via deterministic flow matching from a primary noise sample and a perturbed copy; the per-step L2 divergence δ_i between the two rollouts (Eq. 2) is used as a proxy for prediction instability. A dual-threshold rule (absolute threshold θ_abs and ratio threshold r_thresh, Eqs. 3-5) truncates execution at the first 'unstable' step, with the executed horizon clamped to [s_min, s_max] (Eq. 6). The method is evaluated on a real da Vinci-like dual-arm platform with 600 teleoperated demonstrations across needle (reach, pick, regrasp) and tissue (reach, lift, resection) suites, comparing against π0.5, π0, RDT, and Octo. The headline results are end-to-end success improvements from 55% to 60% for needle manipulation and from 55% to 80% for tissue manipulation, with an ablation comparing TDHD against fixed-horizon execution strategies.

Significance. If the reported gains are robust, TDHD would offer a lightweight, retraining-free intervention that improves the reliability of vision-language-action policies in contact-rich surgical manipulation, which is a genuine and timely contribution. The strengths of the paper include real-hardware deployment, a new dual-arm surgical benchmark with synchronized multi-view perception, a relatively large teleoperation dataset, hierarchical success metrics, and comparison against several recent VLA baselines. The central idea of using divergence between two flow-matching rollouts as a stability signal is interesting and clearly motivated by local Lyapunov exponents. However, the significance is currently limited by evidentiary issues: the proxy is not validated, the free thresholds are unreported, the main success differences are within the noise of 20-trial evaluations, and the ablation lacks a matched random-truncation control. These issues do not invalidate the concept, but they do mean the quantitative claims are not yet established.

major comments (4)
  1. [III-C, Eq. (2)] The central claim rests on the assertion that the step-wise L2 divergence δ_i between two flow-matching trajectories generated from z1 and z1 + εη is a reliable proxy for prediction instability and execution risk. This assertion is never tested: no analysis correlates δ_i with measured per-step action error, stage-level failure, or end-to-end success. Moreover, δ_i is computed from a single random perturbation η, making it a high-variance one-sample estimate of sensitivity to initial noise rather than a calibrated uncertainty measure. A validation experiment is needed, e.g., correlating δ_i with downstream position error or with success under different truncation rules, to show that the divergence signal rather than the act of replanning drives the gains.
  2. [IV-E, Table II; IV-D, Table I] The method's free parameters are never reported: θ_abs, r_thresh, ε, s_min, s_max, and the action chunk length H are all absent, and no sensitivity analysis is provided. Because the dual-threshold rule is defined by these parameters, the reader cannot assess how robust the reported gains are to threshold choice. Additionally, the 'Fixed s_best' baseline in Table II is selected by scanning s ∈ {5,10,15,20,25} on what appears to be the evaluation set, which is selection on the test data and invalidates it as a fair comparison. Finally, the ablation text states there were 'minor setup changes (e.g., changes in camera viewpoints),' so Table II is not directly comparable to Table I (for example, fixed s_max tissue reach drops from 80% in Table I to 70% in Table II). These issues leave room for the observed gains to reflect hidden calibration rather than the divergence mechanism.
  3. [IV-D, Table I] No confidence intervals or significance tests are reported for any success rate. With 20 trials per task setting, the headline needle improvement from 55% to 60% is a difference of one successful trial (11/20 vs 12/20), and the pick conditional improvement from 85% to 90% is also one trial (17/20 vs 18/20). Under an exact binomial test, the needle total difference is far from significant. The claim in the abstract and Section IV-D that TDHD 'consistently improves performance' is not supported by the needle result without additional statistical evidence or a larger number of trials. At minimum, the paper should report exact binomial confidence intervals and pairwise tests, and ideally increase the trial count for the headline conditions.
  4. [IV-E] The ablation cannot separate the mechanism's specific benefit from a general benefit of more frequent replanning during contact-sensitive stages. The comparison is only against fixed-horizon execution; there is no control that truncates at matched random time steps or at fixed intervals calibrated to TDHD's average executed horizon (e.g., 17-23 steps). Since the paper reports that TDHD tends to truncate during contact transitions and the largest gains appear in exactly those stages, the observed improvement could be explained by 'replan more often in hard stages' rather than by 'divergence correctly identifies hard steps.' Adding a random-truncation control with matched replanning frequency would directly address this underdetermination.
minor comments (5)
  1. [IV-B] The dataset description is ambiguous: it says 'we collected exactly 100 successful trials' and then 'we allocated 80 trials for the training set and 20 for testing.' It is unclear whether the 20 test trials are successful demonstrations used for evaluation or whether the 20 trials per policy in Table I are independent re-executions that may include failures. Please clarify the relationship between the demonstration split and the evaluation protocol.
  2. [IV-E] The ablation reports only aggregate success rates over 10 runs per condition; no variance or per-run breakdown is given. Given the small sample, reporting exact counts or confidence intervals would help interpret the differences (e.g., 90% vs 80% in needle pick could be 9/10 vs 8/10).
  3. [IV-E, final paragraph] The sentence listing average executed step lengths for needle manipulation (21.15, 22.08, 20.40) appears twice, once in the paragraph preceding the analysis and again in the analysis itself. Please remove the duplicate.
  4. [III-C, Eq. (4)] The ratio threshold uses ξ in the denominator but its value is never specified; even a small positive constant should be reported, since the ratio rule can be sensitive to ξ when δ_{i-1} is near zero.
  5. [V] The conclusion states that future work will investigate 'threshold calibration, the impact of trajectory ensemble size in TDHD, and its consistency across alternative VLA models.' These are appropriate future directions, but the first two items directly bear on the validity of the current results and should be addressed at least partially in this paper, not deferred entirely, given that thresholds are unreported.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: TDHD's divergence proxy is an untested empirical assumption, and threshold tuning is a reproducibility concern, but no prediction reduces to a fitted input or to a self-citation chain.

full rationale

The derivation chain is not circular. TDHD defines per-step divergence δ_i as the L2 distance between two flow-matching-generated trajectories from z1 and z1+εη (Eq. 2) and truncates execution via absolute and ratio thresholds (Eqs. 3–6). The headline success rates (55%→60% needle, 55%→80% tissue) are measured on real hardware and are not algebraically equivalent to δ_i; there is no equation-level reduction of the reported gains to the threshold rule. The assumption that δ_i tracks execution risk is an untested empirical hypothesis, and Section V's statement that 'Future work will investigate threshold calibration' admits the thresholds are not systematically calibrated; these are correctness/reproducibility concerns, not circularity. The ablation's 'Fixed s_best selected from s∈{5,10,15,20,25}' is a baseline constructed by selection, which could weaken the comparison if selection used the evaluation set, but it does not make the TDHD prediction equivalent to its input. The paper cites the authors' own prior work ([13], [21], [32]) only in related-work contexts; none of these citations carries the load of the central claim, and no uniqueness theorem or prior result by the same authors is invoked to force the method. The central derivation is therefore self-contained and benchmarked against external hardware results.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim depends on unreported threshold parameters and on an unvalidated mapping from trajectory divergence to execution unreliability. The empirical evaluation introduces additional assumptions about demonstration quality and trial-count adequacy. No new physical entities are introduced; TDHD is a control-layer algorithm.

free parameters (5)
  • Absolute divergence threshold theta_abs = not reported
    Dual-threshold stopping rule in Eq. (3); value chosen by the authors but never given.
  • Ratio threshold r_thresh = not reported
    Jump-detection stopping rule in Eq. (4); value not reported.
  • Noise perturbation scale epsilon = not reported
    Constructs the perturbed noise z_2 = z_1 + epsilon*eta; magnitude not specified.
  • Execution horizon bounds s_min and s_max = not reported (s_max = 25 or 15 used in ablation)
    The final horizon s* is clamped to [s_min, s_max] in Eq. (6); the paper gives the ablation values but not the TDHD settings.
  • Action chunk length H = not reported
    The length of the predicted action chunk affects divergence computation and replanning cadence; not specified.
assumptions (4)
  • standard math Flow-matching ODE output is a deterministic function of the initial noise z for fixed conditioning and solver settings.
    Needed so the two rollouts differ only via the noise perturbation, as stated around Eq. (1).
  • ad hoc to paper L2 divergence between two perturbed rollouts is a valid proxy for step-wise action reliability.
    The core premise of TDHD, asserted in Section III-C without independent validation against measured execution errors.
  • ad hoc to paper Dual thresholds can be set so that truncation improves success without excessive replanning or premature stopping.
    The stopping rule in Eqs. (3)-(5) is assumed effective, but the threshold values are unreported and their tuning is not analyzed.
  • domain assumption Teleoperated demonstrations at 30 Hz with NTP synchronization are sufficient for training VLA policies on these six subtasks.
    Dataset construction in Section IV-B assumes expert teleoperation provides adequate coverage and label quality for the target subtasks.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Trajectory Divergence Horizon Decision for Reliable Dual-Arm Surgical Subtask Manipulation." pith.science (2026). https://pith.science/paper/UZB52A5S

@misc{pith2026260809125,
  author       = {Pith},
  title        = {Pith review of: Trajectory Divergence Horizon Decision for Reliable Dual-Arm Surgical Subtask Manipulation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UZB52A5S}},
  note         = {Machine review of arXiv:2608.09125}
}
read the original abstract

Surgical robotic systems are increasingly being adopted as clinical workload rises, motivating autonomous solutions for repetitive manipulation subtasks. Learning-based controllers improve generalization compared with rule-based and analytic approaches, but most are trained for individual tasks and remain difficult to reuse across procedures. Vision-Language-Action (VLA) models provide a unified framework that integrates visual perception, language grounding, and action generation, offering a promising path toward more composable surgical autonomy. However, existing VLA policies rely on fixed-length open-loop action sequences, where changing scene conditions can lead to accumulated errors and potential risks in surgical manipulation. To mitigate this issue, we formulate surgical VLA deployment as an adaptive execution-horizon decision problem and propose Trajectory Divergence Horizon Decision (TDHD), a test-time mechanism that estimates step-wise action reliability by measuring the divergence between two flow-matching-generated trajectories under small noise perturbations and truncates execution using a dual-threshold rule to trigger timely replanning. We further establish a real-world da Vinci-like dual-arm benchmark with synchronized multi-view perception and language instructions, and collect 600 teleoperated demonstrations across needle (reach, pick, regrasp) and tissue (reach, lift, resection) manipulation suites. On real hardware with 20 trials per task setting, TDHD consistently improves performance over the latest VLA baselines: success increases from 55\% to 60\% for needle manipulation and from 55\% to 80\% for tissue manipulation, with the largest gains observed in the final manipulation stages. These results highlight the importance of adaptive execution control for reliable deployment of VLA models in surgical robotic manipulation.

Figures

Figures reproduced from arXiv: 2608.09125 by the authors.

Figure 1
Figure 1. Overview of the automated surgical manipulation framework with TDHD. (A) Multimodal context information includes multi-view images, a language instruction, and robotic arm proprioception; (B) VLA prefix construction, where visual, linguistic, and proprioceptive tokens are fused by a VLM Transformer into a Prefix KV Cache computed once per cycle; (C) Noise sampling creates a primary noise z1 and a perturbed copy z2=z… view at source ↗
Figure 2
Figure 2. Experimental setup of the dual-arm da Vinci-like surgical robot. The platform integrates multi-view perception from exterior and wrist￾mounted cameras and collects teleoperated demonstrations by controlling the dVRK via Geomagic Touch controllers. camera for global RGB-D sensing and two wrist cameras for local RGB views, as shown in [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Multi-view camera observations during real-world deployment. Each column shows synchronized frames from three cameras (exterior, left￾wrist, right-wrist) at the same timestamp. The horizontal axis represents the temporal progression of keyframes across needle manipulation (top) and tissue manipulation (bottom) task suites, illustrating the sequential reach, pick/lift, and regrasp/resection stages. TABLE I QUANTITATI… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 24 canonical work pages

  1. [1]

    A decade retrospective of medical robotics research from 2010 to 2020,

    P. E. Dupont, B. J. Nelson, M. Goldfarb, B. Hannaford, A. Menciassi, M. K. O’Malley, N. Simaan, P. Valdastri, and G.-Z. Yang, “A decade retrospective of medical robotics research from 2010 to 2020,”Science Robotics, vol. 6, no. 60, p. eabi8017, 2021

    work page 2010

  2. [2]

    Surgical embodied intelligence for generalized task autonomy in laparoscopic robot-assisted surgery,

    Y . Long, A. Lin, D. H. C. Kwok, L. Zhang, Z. Yang, K. Shi, L. Song, J. Fu, H. Lin, W. Weiet al., “Surgical embodied intelligence for generalized task autonomy in laparoscopic robot-assisted surgery,” Science Robotics, vol. 10, no. 104, p. eadt3093, 2025

    work page 2025

  3. [3]

    Autonomous robotic laparoscopic surgery for intestinal anastomosis,

    H. Saeidi, J. D. Opfermann, M. Kam, S. Wei, S. L ´eonard, M. H. Hsieh, J. U. Kang, and A. Krieger, “Autonomous robotic laparoscopic surgery for intestinal anastomosis,”Science Robotics, vol. 7, no. 62, p. eabj2908, 2022

    work page 2022

  4. [4]

    Learning by observation for surgical subtasks: Multilateral cutting of 3d viscoelastic and 2d orthotropic tissue phantoms,

    A. Murali, S. Sen, B. Kehoe, A. Garg, S. McFarland, S. Patil, W. D. Boyd, S. Lim, P. Abbeel, and K. Goldberg, “Learning by observation for surgical subtasks: Multilateral cutting of 3d viscoelastic and 2d orthotropic tissue phantoms,” in2015 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2015, pp. 1202–1209

    work page 2015

  5. [5]

    Real-time glottis detection framework via spatial-decoupled feature learning for nasal transnasal intubation,

    J. Liu, G. Zhang, Y . Zhou, R. Hao, Y . Zhang, and H. Ren, “Real-time glottis detection framework via spatial-decoupled feature learning for nasal transnasal intubation,”Biomimetic Intelligence and Robotics, p. 100325, 2026

    work page 2026

  6. [6]

    Deep learning–based autonomous retinal vein cannulation in ex vivo porcine eyes,

    P. Zhang, P. Gehlbach, R. H. Taylor, I. Iordachita, and M. Kobilarov, “Deep learning–based autonomous retinal vein cannulation in ex vivo porcine eyes,”Science Robotics, vol. 10, no. 109, p. eadw2969, 2025

    work page 2025

  7. [7]

    Vision language action models in robotic manipulation: A systematic review,

    M. U. Din, W. Akram, L. S. Saoud, J. Rosell, and I. Hussain, “Vision language action models in robotic manipulation: A systematic review,” arXiv preprint arXiv:2507.10672, 2025

    arXiv 2025

  8. [8]

    General-purpose foundation models for increased autonomy in robot- assisted surgery,

    S. Schmidgall, J. W. Kim, A. Kuntz, A. E. Ghazi, and A. Krieger, “General-purpose foundation models for increased autonomy in robot- assisted surgery,”Nature Machine Intelligence, vol. 6, no. 11, pp. 1275–1283, 2024

    work page 2024

Show all 38 references
  1. [9]

    Vaisi: Vision-based adaptive impedance-control for surgical incisions,

    J. Lee, T. Marotta, S. McLachlin, A. Wong, and Y . Hu, “Vaisi: Vision-based adaptive impedance-control for surgical incisions,” in 2025 IEEE-RAS 24th International Conference on Humanoid Robots (Humanoids). IEEE, 2025, pp. 1–8

    2025

  2. [10]

    Task-adaptive analytical affordance estimation for feature-based manipulation of soft tissues in robotic surgery,

    S. Yang, Y . Wang, Y . Wang, H. Zhao, and H. Ding, “Task-adaptive analytical affordance estimation for feature-based manipulation of soft tissues in robotic surgery,”IEEE Transactions on Automation Science and Engineering, vol. 23, pp. 1301–1314, 2025

    2025

  3. [11]

    Artificial in- telligence: revolutionizing robotic surgery,

    M. Iftikhar, M. Saqib, M. Zareen, and H. Mumtaz, “Artificial in- telligence: revolutionizing robotic surgery,”Annals of Medicine and Surgery, vol. 86, no. 9, pp. 5401–5409, 2024

    2024

  4. [12]

    Sensorless planning for medical needle insertion procedures,

    R. Alterovitz, K. Goldberg, J. Pouliot, R. Taschereau, and I.-C. Hsu, “Sensorless planning for medical needle insertion procedures,” in Proceedings 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2003), vol. 4. IEEE, 2003, pp. 3337–3343

    2003

  5. [13]

    Geolang: Geometry-aware language-guided grasping with unified rgb-d multimodal learning,

    R. Tang, G. Wang, L. Bai, H. Gao, J. Lai, C. K. Ng, J. Wang, F. Zhang, and H. Ren, “Geolang: Geometry-aware language-guided grasping with unified rgb-d multimodal learning,”arXiv preprint arXiv:2602.04231, 2026

    2026

  6. [14]

    Mosformer: Momentum encoder-based inter-slice fusion transformer for medical image segmentation,

    D. Huang, X. Zhou, M. Gui, X. Xie, S. Liu, S. Wang, Z. Feng, Z. Lai, and Z. Hou, “Mosformer: Momentum encoder-based inter-slice fusion transformer for medical image segmentation,”Biomimetic Intelligence and Robotics, vol. 6, no. 2, p. 100298, 2026

    2026

  7. [15]

    A user-centered shared control scheme with learning from demonstration for robotic surgery,

    H. Zheng, Z. J. Hu, Y . Huang, X. Cheng, Z. Wang, and E. Burdet, “A user-centered shared control scheme with learning from demonstration for robotic surgery,” in2024 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2024, pp. 15 195–15 201

    2024

  8. [16]

    Autonomous soft tissue retraction using demonstration-guided reinforcement learning,

    A. Singh, W. Shi, and M. D. Wang, “Autonomous soft tissue retraction using demonstration-guided reinforcement learning,”arXiv preprint arXiv:2309.00837, 2023

    2023 arXiv

  9. [17]

    Surrol: An open-source reinforcement learning centered and dvrk compatible platform for surgical robot learning,

    J. Xu, B. Li, B. Lu, Y .-H. Liu, Q. Dou, and P.-A. Heng, “Surrol: An open-source reinforcement learning centered and dvrk compatible platform for surgical robot learning,” in2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2021, pp. 1821–1828

    2021

  10. [18]

    Surgical robot transformer (srt): Imitation learning for surgical tasks,

    J. W. Kim, T. Z. Zhao, S. Schmidgall, A. Deguet, M. Kobilarov, C. Finn, and A. Krieger, “Surgical robot transformer (srt): Imitation learning for surgical tasks,”arXiv preprint arXiv:2407.12998, 2024

    2024 arXiv

  11. [19]

    Moe-act: Improving surgical imitation learning policies through supervised mixture-of- experts,

    L. Mazza, A. Rodriguez, R. Younis, M. Lelis, O. Hellig, C. Li, S. Bodenstedt, M. Wagner, and S. Speidel, “Moe-act: Improving surgical imitation learning policies through supervised mixture-of- experts,”arXiv preprint arXiv:2601.21971, 2026

    2026 arXiv

  12. [20]

    Sufia-bc: Generating high quality demonstration data for visuomotor policy learning in surgical subtasks,

    M. Moghani, N. Nelson, M. Ghanem, A. Diaz-Pinto, K. Hari, M. Az- izian, K. Goldberg, S. Huver, and A. Garg, “Sufia-bc: Generating high quality demonstration data for visuomotor policy learning in surgical subtasks,” in2025 IEEE International Conference on Robotics and Automati...

    2025

  13. [21]

    Copesd: A multi-level surgical motion dataset for train- ing large vision-language models to co-pilot endoscopic submucosal dissection,

    G. Wang, H. Xiao, R. Zhang, H. Gao, L. Bai, X. Yang, Z. Li, H. Li, and H. Ren, “Copesd: A multi-level surgical motion dataset for train- ing large vision-language models to co-pilot endoscopic submucosal dissection,” inProceedings of the 33rd ACM International Conference on Mu...

    2025

  14. [22]

    Srt-h: A hierarchical framework for autonomous surgery via language-conditioned imitation learning,

    J. W. Kim, J.-T. Chen, P. Hansen, L. X. Shi, A. Goldenberg, S. Schmidgall, P. M. Scheikl, A. Deguet, B. M. White, D. R. Tsai et al., “Srt-h: A hierarchical framework for autonomous surgery via language-conditioned imitation learning,”Science Robotics, vol. 10, no. 104, p. eadt...

    2025

  15. [23]

    Vision- language-action (vla) models: Concepts, progress, applications and challenges,

    R. Sapkota, Y . Cao, K. I. Roumeliotis, and M. Karkee, “Vision- language-action (vla) models: Concepts, progress, applications and challenges,”arXiv preprint arXiv:2505.04769, 2025

    2025

  16. [24]

    Rt-2: Vision-language-action models transfer web knowledge to robotic control,

    B. Zitkovich, T. Yu, S. Xu, P. Xu, T. Xiao, F. Xia, J. Wu, P. Wohlhart, S. Welker, A. Wahidet al., “Rt-2: Vision-language-action models transfer web knowledge to robotic control,” inConference on Robot Learning. PMLR, 2023, pp. 2165–2183

    2023

  17. [25]

    Octo: An open-source generalist robot policy,

    O. M. Team, D. Ghosh, H. Walke, K. Pertsch, K. Black, O. Mees, S. Dasari, J. Hejna, T. Kreiman, C. Xuet al., “Octo: An open-source generalist robot policy,”arXiv preprint arXiv:2405.12213, 2024

    2024 arXiv

  18. [26]

    Open- vla: An open-source vision-language-action model,

    M. J. Kim, K. Pertsch, S. Karamcheti, T. Xiao, A. Balakrishna, S. Nair, R. Rafailov, E. Foster, G. Lam, P. Sanketiet al., “Open- vla: An open-source vision-language-action model,”arXiv preprint arXiv:2406.09246, 2024

    2024 arXiv

  19. [27]

    π 0: A vision- language-action flow model for general robot control,

    K. Black, N. Brown, D. Driess, A. Esmail, M. Equi, C. Finn, N. Fusai, L. Groom, K. Hausman, B. Ichteret al., “π 0: A vision- language-action flow model for general robot control,”arXiv preprint arXiv:2410.24164, 2024

    2024 arXiv

  20. [28]

    π 0.5: a vision-language-action model with open-world generalization,

    P. Intelligence, K. Black, N. Brown, J. Darpinian, K. Dhabalia, D. Driess, A. Esmail, M. Equi, C. Finn, N. Fusaiet al., “π 0.5: a vision-language-action model with open-world generalization,”arXiv preprint arXiv:2504.16054, 2025

    2025 arXiv

  21. [29]

    Galaxea open-world dataset and g0 dual-system vla model,

    T. Jiang, T. Yuan, Y . Liu, C. Lu, J. Cui, X. Liu, S. Cheng, J. Gao, H. Xu, and H. Zhao, “Galaxea open-world dataset and g0 dual-system vla model,”arXiv preprint arXiv:2509.00576, 2025

    2025 arXiv

  22. [30]

    Rdt-1b: a diffusion foundation model for bimanual manipulation,

    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

    2024 arXiv

  23. [31]

    Robonurse- vla: Robotic scrub nurse system based on vision-language-action model,

    S. Li, J. Wang, R. Dai, W. Ma, W. Y . Ng, Y . Hu, and Z. Li, “Robonurse- vla: Robotic scrub nurse system based on vision-language-action model,” in2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). IEEE, 2025, pp. 3986–3993

    2025

  24. [32]

    Endovla: Dual-phase vision-language-action for precise autonomous tracking in endoscopy,

    C. K. Ng, L. Bai, G. Wang, Y . Wang, H. Gao, C. Jin, T. Zeng, H. Renet al., “Endovla: Dual-phase vision-language-action for precise autonomous tracking in endoscopy,” in9th Annual Conference on Robot Learning, 2025

    2025

  25. [33]

    Local lyapunov exponents in chaotic systems,

    B. Eckhardt and D. Yao, “Local lyapunov exponents in chaotic systems,”Physica D: Nonlinear Phenomena, vol. 65, no. 1-2, pp. 100– 108, 1993

    1993

  26. [34]

    Do what you say: Steering vision-language-action models via runtime reasoning-action alignment verification,

    Y . Wu, A. Li, T. Hermans, F. Ramos, A. Bajcsy, and C. P ˜AˇSrez- D’Arpino, “Do what you say: Steering vision-language-action models via runtime reasoning-action alignment verification,”arXiv preprint arXiv:2510.16281, 2025

    2025

  27. [35]

    Deep rein- forcement learning in a handful of trials using probabilistic dynamics models,

    K. Chua, R. Calandra, R. McAllister, and S. Levine, “Deep rein- forcement learning in a handful of trials using probabilistic dynamics models,”Advances in neural information processing systems, vol. 31, 2018

    2018

  28. [36]

    The intuitive/sup tm/telesurgery system: overview and application,

    G. S. Guthart and J. K. Salisbury, “The intuitive/sup tm/telesurgery system: overview and application,” inProceedings 2000 ICRA. Mil- lennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings (Cat. No. 00CH37065), vol. 1. IEEE, 2000, pp...

    2000

  29. [37]

    1405 telelap alf-x: a novel telesurgical system for the 21st century,

    E. Altobelli, S. Gidaro, A. M. Bove, C. Falavolti, E. Ruiz, M. Stark, and M. Buscarini, “1405 telelap alf-x: a novel telesurgical system for the 21st century,”The Journal of Urology, vol. 189, no. 4S, pp. e575– e576, 2013

    2013

  30. [38]

    Dlr mirosurge: a versatile system for research in endoscopic telesurgery,

    U. Hagn, R. Konietschke, A. Tobergte, M. Nickl, S. J ¨org, B. K ¨ubler, G. Passig, M. Gr ¨oger, F. Fr¨ohlich, U. Seiboldet al., “Dlr mirosurge: a versatile system for research in endoscopic telesurgery,”International journal of computer assisted radiology and surgery, vol. 5, ...

    2010

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.