Pith. sign in

REVIEW 4 major objections 5 minor 31 references

HMVLA: Hyperbolic Multimodal Fusion for Vision-Language-Action Models

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read HMVLA claims that embedding image and text features in hyperbolic space while fusing them through a sparsely gated mixture of experts improves vision-language-action control accuracy and cross-domain generalization.

desk verdict Plausible architecture, thin evidence: the hyperbolic + MoE fusion story is coherent, but the key generalization experiment rests on an undescribed dataset and there are no seeds or error bars. read the letter →

arxiv 2602.02533 v1 pith:25ZPKNVJ submitted 2026-01-28 cs.RO cs.LG

Kun Wang , Xiao Feng , Mingcheng Qu , Tonghua Su This is my paper · ORCID
classification cs.ROcs.LG
keywords vision-language-actionmodelshyperbolicspaceLorentzmodelentailmentlossmixtureofexpertssemanticalignmentLIBERObenchmarkrobotcontrol
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

The paper sets out to show that the usual practice of feeding contrastive visual and linguistic features directly into a robot policy loses the hierarchical structure that connects an instruction like 'pick up the black bowl' to its visual referents, and that this loss is fixable by geometry. HMVLA projects image and text embeddings onto a Lorentz hyperboloid, the standard model of hyperbolic space, and adds an entailment-cone loss that penalizes an image embedding falling outside the cone of its paired text; a soft mixture-of-experts module then fuses the modalities through weighted expert routing with load balancing. Built on the Dita diffusion-transformer policy, HMVLA reports 86% average task accuracy on LIBERO versus 82% for Dita and 81% for CoT-VLA, and describes its gains on a self-constructed generalization dataset (Gen) as 'breakthrough improvements.' The cared-about consequence is that hyperbolic alignment is a practical, drop-in component for instruction-following robot policies, improving both task accuracy and generalization to novel objects and instructions.

What carries the argument

The load-bearing object is the Lorentz model of hyperbolic space, the upper sheet of a two-sheeted hyperboloid in $\mathbb{R}^{n+1}$ with Lorentzian inner product $\langle x,y\rangle_L = \langle x_{\mathrm{space}},y_{\mathrm{space}}\rangle - x_{\mathrm{time}}y_{\mathrm{time}}$; its exponential expansion makes it a natural home for hierarchical, tree-like data. The paper parameterizes only the spatial components at the origin, so the exponential map reduces to a scaling of the Euclidean embedding by $\sinh(\sqrt{c}\|v\|)/(\sqrt{c}\|v\|)$, then defines an entailment cone with half-aperture $\operatorname{aper}(x)=\sin^{-1}(2K/(\sqrt{c}\|x_{\mathrm{space}}\|))$ and a margin loss $L_{\mathrm{ent}}=\max(0,\operatorname{ext}(x,y)-\operatorname{aper}(x))$ that punishes image embeddings outside their paired text's cone. The second mechanism is a soft mixture-of-experts module inserted into the Q-Former: each query token gets softmax gating weights and is updated as a weighted sum of expert outputs, with a load-balancing loss added to the task loss. These two pieces carry the claimed improvement: the cone encodes vision-language hierarchy, and the router decomposes fine-grained instruction semantics across experts.

What would settle it

Run HMVLA and Dita on the four public LIBERO splits for five independent seeds under identical compute, and separately re-create the Gen split from a fully specified held-out task list; if the average accuracy gap does not reach the reported 4 points or the Gen delta is not reproducible, the central claim is falsified. A second quick check is to set $\lambda = 0$ (no entailment loss) or disable the MoE and retrain with the same budget; if either removal leaves accuracy unchanged, that component is not load-bearing.

Watch

Extended reading notes

Core claim

The paper's central claim is that Euclidean contrastive alignment, the standard way of tying image and text embeddings, is not the right interface for vision-language-action models, because it distorts the tree-like hierarchy of semantics: a model can learn spurious correlations such as 'white' with 'background' instead of grounding 'grasp' in the cup. HMVLA claims to preserve hierarchy by mapping CLIP image and text embeddings to a Lorentz hyperboloid, using the simplified exponential map at the origin, and by enforcing an entailment-cone constraint with loss $L_{\mathrm{ent}}$ on the exterior angle between paired embeddings. A soft mixture-of-experts module fused into the Q-Former routes each fused query token across experts through softmax weights, replacing the standard feed-forward layer, and a load-balancing loss keeps expert usage even. The paper reports that this combination reaches 90/96/89/69 percent on the Spatial, Object, Goal, and LONG LIBERO splits (86 average) against Dita's 84/96/85/63 (82 average), and shows larger margins on its reconstructed Gen dataset, with ablations attributing the gain to both the hyperbolic loss and the mixture-of-experts module.

Load-bearing premise

The load-bearing premise is that the reconstructed Gen dataset is a valid cross-domain generalization test of novel objects and instructions; if that dataset was built in a way that resembles the training splits or leaks task information, the reported generalization advantage would evaporate, leaving only the in-distribution LIBERO gains.

Editorial extensions

If this is right

  • If the entailment cone is doing the work the paper claims, then instructions with compositional attributes like 'black bowl' or 'green triangle' should become more robust to distractors, because attribute structure is preserved rather than flattened into a single vector.
  • Because the MoE slot is a drop-in replacement for the feed-forward layer in the Q-Former and the hyperbolic projection sits before fusion, the same recipe can be applied to other VLA backbones without changing their diffusion action decoding.
  • The reported 4-point average gain over Dita on LIBERO, if it holds across seeds, implies that Euclidean alignment is a measurable bottleneck even on in-distribution benchmark tasks, not only on novel-object generalization.
  • A correct HMVLA would provide a diagnostic handle: expert routing weights can be inspected to see whether color, shape, and location components of an instruction are being handled by distinct experts, which is exactly the decomposition the paper claims the MoE gives.

Reading between the lines

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

  • The authors' ablation compares Dita with hyperbolic loss and with MoE separately, but they never combine exactly one mechanism at a time against the full model under matched seeds; a head-to-head of Dita plus hyperbolic-only versus Dita plus MoE-only would isolate which component carries the LIBERO gain and which carries the Gen gain.
  • The routing weights of the soft MoE could be visualized per attribute word; if color, shape, and location cluster onto distinct experts, the experts could be reused as a zero-shot skill library for novel task instructions, an extension the paper does not explore.
  • The entailment-cone loss is applied only to image-text pairs; applying the same hierarchical penalty to latent action tokens or video frames would test whether the hierarchy claim extends from perception to control, which the paper leaves open.
  • The Gen dataset is the load-bearing piece of the generalization story, yet its construction protocol is absent; until a public protocol is given, the 'breakthrough' gap cannot be compared across papers or re-run by other labs, so the strength of the generalization claim is currently an editorial unknown.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The manuscript proposes HMVLA, a vision-language-action model built on the Dita backbone. The method embeds image and text features in the Lorentz model of hyperbolic space, adds an entailment-cone loss to the contrastive objective (Eqs. 11-14), and replaces the Q-Former feed-forward layers with a soft mixture-of-experts module plus a load-balancing loss (Eqs. 15-18). The authors evaluate on four LIBERO datasets and on a self-constructed dataset called Gen, reporting that HMVLA achieves 86% average task accuracy versus 82% for Dita (Table 1), larger gains on Gen (Fig. 3), and component-level improvements in an ablation (Fig. 5). The central claim is that hyperbolic semantic alignment and soft MoE fusion improve both task accuracy and cross-domain generalization.

Significance. If the empirical results are reproducible, the paper would make a useful contribution: applying hyperbolic geometry to VLA is a novel and plausible direction, and the soft-MoE fusion mechanism is a reasonable architectural choice. The use of the external LIBERO benchmark and comparisons to several recent baselines are appropriate. However, as presented, the evidence is not yet sufficient to validate the central claims. The Gen dataset is not auditable, no statistical precision is reported for any result, and key baseline and implementation details are missing. The method section is largely coherent and the direction is worth pursuing, but the empirical core needs substantial strengthening.

major comments (4)
  1. [Section 3 (Datasets; Fig. 3)] The construction of the Gen dataset is described only as 'reconstructed a new dataset (Gen) for training and validation'; no task list, object set, instruction-template design, demo count, background or texture randomization protocol, or train/validation split rule is provided. Because Fig. 3's claimed 'breakthrough improvements' on Gen are the sole evidence for the cross-domain generalization claim, this experiment is unauditable. The authors must either publish the full dataset-construction protocol and the data, or replace the generalization evaluation with a publicly available benchmark, before the generalization claim can be assessed.
  2. [Table 1; Fig. 5] All accuracy numbers appear to be single-run values, with no standard deviation, no number of seeds, and no statistical test. The headline gap of 86% versus 82% over Dita is small enough that seed-to-seed variance in diffusion-policy training on LIBERO could plausibly affect it. Please report mean and standard deviation over at least five seeds for the proposed method and all baselines, and state whether the baseline numbers are taken from prior papers or obtained by retraining under identical schedules, resolutions, and step counts.
  3. [Section 2.1, Eqs. (11)-(13)] The entailment-cone loss is central to the claimed semantic-alignment benefit, but its specification is incomplete. The aperture depends on a boundary constant K whose value is never reported, the curvature c and the weighting lambda are fixed without sensitivity analysis, and Eq. (12) is given without derivation. Please provide the precise cone construction, the value of K, and an ablation over c and lambda to demonstrate that the loss, rather than the extra parameters, is responsible for any improvement.
  4. [Section 3; Fig. 3] Figure 3 has no printed numerical values, so the reader cannot determine the exact Gen accuracies or the magnitudes of the claimed improvements. Additionally, the figure legend and text do not specify whether OpenVLA, DFP-OTTER, and OTTER were run on the same Gen dataset under the same protocol; without this information, the comparison is not interpretable. Please add exact numbers and a complete experimental protocol.
minor comments (5)
  1. [Section 3, first paragraph after Table 1] The sentence 'As shown in Table 1 and Fig. 3, we conducted comparative experiments...' is repeated verbatim in consecutive paragraphs; please remove the duplicate.
  2. [Eqs. (5)-(10)] The exponential-map notation is inconsistent: Eq. (5) writes expm_z(v) but the subscript is not used in the formula, and Eqs. (7)-(8) suppress the time coordinate by writing cosh(...)0 instead of the origin's time component. Please define all symbols and clearly separate the spatial and time components.
  3. [Eq. (17)] The load-balancing loss uses n_m as 'the number of tokens routed to expert m', but the routing weights are soft and every token contributes to every expert; please clarify how n_m is computed from the soft weights.
  4. [Fig. 5] The ablation figure has no numeric labels on the bars, and the horizontal axis mixes 'Long' with the dataset name 'LONG' used in Table 1. Please add the values and standardize the naming.
  5. [Section 2.2] The text says the MoE replaces feed-forward layers but does not state the number of experts used in the main experiments beyond the implementation-details sentence (6 experts); please state whether the same 6 experts are used in all ablations and what happens if the number of experts is varied.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: benchmark numbers are external, losses are imposed, and no self-citation is load-bearing.

full rationale

The paper's derivation chain is not circular. HMVLA's objective (Eq. 14, L = L_cont + λL_ent) is an imposed training loss, not a quantity recovered from the benchmark outcome; c=0.1 and λ=0.1 are ordinary hyperparameter choices, and the entailment-cone/MoE components are evaluated as interventions on the external Dita backbone. The LIBERO results in Table 1 are measured against a standard benchmark, and the comparison to Dita [21] is an external baseline, not a self-citation. No load-bearing claim is justified solely by the authors' own prior work, and no uniqueness theorem or ansatz is smuggled in via citation. The reconstructed Gen dataset is described only as 'reconstructed a new dataset (Gen) for training and validation' with no construction protocol, which is an auditability/reproducibility weakness rather than a circular step: nothing in the text shows that the reported Gen gains are equivalent by construction to the model's inputs or to any fitted parameter. Thus the central accuracy and generalization claims retain independent empirical content, and the paper warrants a circularity score of 0.

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

The ledger shows that the paper's novelty is architectural rather than conceptual: it transplants a known hyperbolic entailment loss and a known MoE routing mechanism into a VLA policy backbone, while several hyperparameters that the central claim depends on are either unreported or untested.

free parameters (5)
  • hyperbolic curvature c = 0.1
    Set in Implementation Details; controls geometry, exponential map, and the aperture of the entailment cone. No sensitivity study is reported.
  • entailment loss weight lambda = 0.1
    Balancing coefficient in Eq. 14; chosen without ablation or search.
  • entailment cone boundary constant K = not specified
    Introduced in Eq. 11 as 'a constant K representing boundary conditions near the origin'; no value or tuning procedure is reported.
  • load-balancing weight beta = not specified
    Appears in Eq. 18; no value is given in Implementation Details.
  • number of MoE experts M = 6
    Set in Implementation Details; no ablation over expert counts is reported.
assumptions (5)
  • domain assumption Vision-language semantics in robot instructions have a hierarchical, tree-like structure that hyperbolic space represents with less distortion than Euclidean space.
    Motivates the whole method in the Introduction and Section 2.1; no quantitative evidence of hierarchy is provided.
  • ad hoc to paper The entailment cone penalty in Eq. 13 contributes to semantic alignment rather than acting as an arbitrary regularizer.
    The loss is stated directly; no derivation or independent validation shows it enforces the claimed hierarchy.
  • domain assumption The reconstructed Gen dataset is a valid test of cross-domain generalization.
    Section 3 describes Gen only as 'reconstructed'; the construction, task distribution, and splits are not detailed.
  • domain assumption Reported baseline numbers (DP, Octo, Tra-MoE, CoT-VLA, Dita) were produced under comparable protocols.
    Table 1 mixes numbers from prior papers with no statement about training data, compute, or evaluation seeds.
  • standard math The Lorentz exponential/log map formulas in Eqs. 7-10 correctly embed the encoder outputs on the hyperboloid.
    Standard hyperbolic geometry, but the paper's presentation contains notational slips that make verification harder.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HMVLA: Hyperbolic Multimodal Fusion for Vision-Language-Action Models." pith.science (2026). https://pith.science/paper/25ZPKNVJ

@misc{pith2026260202533,
  author       = {Pith},
  title        = {Pith review of: HMVLA: Hyperbolic Multimodal Fusion for Vision-Language-Action Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/25ZPKNVJ}},
  note         = {Machine review of arXiv:2602.02533}
}
read the original abstract

Vision Language Action (VLA) models have recently shown great potential in bridging multimodal perception with robotic control. However, existing methods often rely on direct fine-tuning of pre-trained Vision-Language Models (VLMs), feeding semantic and visual features directly into a policy network without fully addressing the unique semantic alignment challenges in the VLA domain. In this paper, we propose HMVLA, a novel VLA framework that exploits the inherent hierarchical structures in vision and language for comprehensive semantic alignment. Unlike traditional methods that perform alignment in Euclidean space, our HMVLA embeds multimodal features in hyperbolic space, enabling more effective modeling of the hierarchical relationships present in image text data. Furthermore, we introduce a sparsely gated Mixture of Experts (MoE) mechanism tailored for semantic alignment, which enhances multimodal comprehension between images and text while improving efficiency. Extensive experiments demonstrate that HMVLA surpasses baseline methods in both accuracy and generalization. In addition, we validate its robustness by reconstructing datasets to further test cross domain adaptability.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

31 extracted references · 22 canonical work pages

  1. [1]

    HMVLA: Hyperbolic Multimodal Fusion for Vision-Language-Action Models

    INTRODUCTION The rapid development of Large-Language-Models (LLMs) and Vision-Language-Models (VLMs) [1, 2, 3, 4] has signif- icantly advanced Vision-Language-Action (VLA) models in robotics [5, 6, 7, 8, 9, 10]. By leveraging the powerful com- prehension capabilities of pre-trained models, VLA systems encode visual and linguistic modalities into feature t...

    work page Pith review arXiv 2026

  2. [2]

    Hyperbolic Semantic Alignment Our HMVLA framework is based on the Lorentz model in hyperbolic geometry, as shown in Figure 2

    METHOD 2.1. Hyperbolic Semantic Alignment Our HMVLA framework is based on the Lorentz model in hyperbolic geometry, as shown in Figure 2. Specifically, the n-dimensional hyperbolic space is represented as the upper sheet of a two-sheeted hyperboloid embedded inR n+1. Fol- lowing the terminology from special relativity [20], we treat the hyperboloid’s axis...

    work page

  3. [3]

    Specifically, we utilized four datasets: Spatial, Object, Goal, and LONG

    EXPERIMENTS Datasets.We used the LIBERO [19] benchmark for evalua- tion. Specifically, we utilized four datasets: Spatial, Object, Goal, and LONG. To validate the generalization capability of our model, we reconstructed a new dataset (Gen) for training and validation to examine its generalizability. Implementation Details.We adopt Dita [21] as the backbon...

    work page

  4. [4]

    By embedding multimodal features into a hyperbolic space, our model effectively cap- tures the inherent hierarchical relationships within image-text data

    CONCLUSION In this paper, we propose HMVLA, a novel Vision-Language- Action (VLA) model that leverages hyperbolic space repre- sentation and a sparsely-gated Mixture-of-Experts (MoE) mechanism to enhance semantic alignment between vision and language for robotic control. By embedding multimodal features into a hyperbolic space, our model effectively cap- ...

    work page

  5. [5]

    62277011), National Key Research and Development Program of China (Grant No

    ACKNOWLEDGEMENT This work was supported by the National Natural Science Foundation of China (Grant No. 62277011), National Key Research and Development Program of China (Grant No. GG-2024-01-02), Project of Chongqing MEITC (Grant No. YJX-2025001001009) and Open Research Fund from Guang- dong Laboratory of Artificial Intelligence and Digital Econ- omy (SZ)...

    work page 2024

  6. [6]

    Gpt-4 technical report,

    Josh Achiam, Steven Adler, Sandhini Agarwal, et al., “Gpt-4 technical report,”arXiv preprint arXiv:2303.08774, 2023

    arXiv 2023

  7. [7]

    The llama 3 herd of models,

    Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, et al., “The llama 3 herd of models,”CoRR, 2024

    work page 2024

  8. [8]

    Llapa: A vision-language model framework for counterfactual-aware procedural planning,

    Shibo Sun, Xue Li, Donglin Di, Mingjie Wei, Lanshun Nie, Wei-Nan Zhang, Dechen Zhan, Yang Song, and Lei Fan, “Llapa: A vision-language model framework for counterfactual-aware procedural planning,” inProceedings of the 33rd ACM International Conference on Multimedia, 2025, pp. 5020–5029

    work page 2025

Show all 31 references
  1. [9]

    Sage: A visual language model for anomaly detection via fact enhancement and entropy-aware alignment,

    Guoxin Zang, Xue Li, Donglin Di, Lanshun Nie, Dechen Zhan, Yang Song, and Lei Fan, “Sage: A visual language model for anomaly detection via fact enhancement and entropy-aware alignment,” inProceedings of the 33rd ACM International Conference on Multimedia, 2025, pp. 5030–5039

    2025

  2. [10]

    Rt- 1: Robotics transformer for real-world control at scale,

    Anthony Brohan, Noah Brown, Justice Carbajal, et al., “Rt- 1: Robotics transformer for real-world control at scale,”arXiv preprint arXiv:2212.06817, 2022

    2022 arXiv

  3. [11]

    Rt-h: Action hierarchies using language,

    Suneel Belkhale, Tianli Ding, Ted Xiao, et al., “Rt-h: Action hierarchies using language,”arXiv preprint arXiv:2403.01823, 2024

    2024 arXiv

  4. [12]

    Vision- language-action model with open-world embodied rea- soning from pretrained knowledge,

    Zhongyi Zhou, Yichen Zhu, Junjie Wen, et al., “Vision- language-action model with open-world embodied rea- soning from pretrained knowledge,”arXiv preprint arXiv:2505.21906, 2025

    2025 arXiv

  5. [13]

    3d-vla: a 3d vision-language-action generative world model,

    Haoyu Zhen, Xiaowen Qiu, Peihao Chen, et al., “3d-vla: a 3d vision-language-action generative world model,” inProceed- ings of the 41st ICML, 2024, pp. 61229–61245

    2024

  6. [14]

    π 0: A Vision-Language-Action Flow Model for General Robot Con- trol,

    Kevin Black, Noah Brown, Danny Driess, et al., “π 0: A Vision-Language-Action Flow Model for General Robot Con- trol,”arXiv preprint arXiv:2410.24164, 2024

    2024 arXiv

  7. [15]

    π 0.5: A Vision-Language-Action Model with Open-World General- ization,

    Physical Intelligence, Kevin Black, Noah Brown, et al., “π 0.5: A Vision-Language-Action Model with Open-World General- ization,”arXiv preprint arXiv:2504.16054, 2025

    2025 arXiv

  8. [16]

    A sur- vey on vision-language-action models for embodied ai,

    Yueen Ma, Zixing Song, Yuzheng Zhuang, et al., “A sur- vey on vision-language-action models for embodied ai,”arXiv preprint arXiv:2405.14093, 2024

    2024 arXiv

  9. [17]

    Vision- language models for vision tasks: A survey,

    Jingyi Zhang, Jiaxing Huang, Sheng Jin, et al., “Vision- language models for vision tasks: A survey,”IEEE TPAMI, vol. 46, no. 8, pp. 5625–5644, 2024

    2024

  10. [18]

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

    Brianna Zitkovich, Tianhe Yu, Sichun Xu, et al., “Rt-2: Vision- language-action models transfer web knowledge to robotic control,” inConference on Robot Learning. PMLR, 2023, pp. 2165–2183

    2023

  11. [19]

    Openvla: An open-source vision-language-action model,

    Moo Jin Kim, Karl Pertsch, Siddharth Karamcheti, et al., “Openvla: An open-source vision-language-action model,” arXiv preprint arXiv:2406.09246, 2024

    2024 arXiv

  12. [20]

    Learn- ing transferable visual models from natural language super- vision,

    Alec Radford, Jong Wook Kim, Chris Hallacy, et al., “Learn- ing transferable visual models from natural language super- vision,” inInternational conference on machine learning. PMLR, 2021, pp. 8748–8763

    2021

  13. [21]

    An ex- tensive study on pre-trained models for program understanding and generation,

    Zhengran Zeng, Hanzhuo Tan, Haotian Zhang, et al., “An ex- tensive study on pre-trained models for program understanding and generation,” inProceedings of the 31st ACM SIGSOFT IS- STA, 2022, pp. 39–51

    2022

  14. [22]

    Hyperbolic spaces,

    Shyan S Chen and Leon Greenberg, “Hyperbolic spaces,” in Contributions to analysis, pp. 49–87. Elsevier, 1974

    1974

  15. [23]

    Hyperbolic image-text representations,

    Karan Desai, Maximilian Nickel, Tanmay Rajpurohit, et al., “Hyperbolic image-text representations,” inInternational Con- ference on Machine Learning. PMLR, 2023, pp. 7694–7731

    2023

  16. [24]

    Libero: Bench- marking knowledge transfer for lifelong robot learning,

    Bo Liu, Yifeng Zhu, Chongkai Gao, et al., “Libero: Bench- marking knowledge transfer for lifelong robot learning,”Ad- vances in Neural Information Processing Systems, vol. 36, pp. 44776–44791, 2023

    2023

  17. [25]

    Zur elektrodynamik bewegter k ¨orper,

    Albert Einstein et al., “Zur elektrodynamik bewegter k ¨orper,” Annalen der physik, vol. 17, no. 10, pp. 891–921, 1905

    1905

  18. [26]

    Dita: Scal- ing diffusion transformer for generalist vision-language-action policy,

    Zhi Hou, Tianyi Zhang, Yuwen Xiong, et al., “Dita: Scal- ing diffusion transformer for generalist vision-language-action policy,” inProceedings of the IEEE/CVF International Con- ference on Computer Vision, 2025

    2025

  19. [27]

    Diffusion policy: Visuomotor policy learning via action diffusion,

    Cheng Chi, Siyuan Feng, Yilun Du, et al., “Diffusion policy: Visuomotor policy learning via action diffusion,” inRobotics: Science and Systems, 2023

    2023

  20. [28]

    Octo: An open-source generalist robot policy,

    Octo Model Team, Dibya Ghosh, Homer Walke, et al., “Octo: An open-source generalist robot policy,”arXiv preprint arXiv:2405.12213, 2024

    2024 arXiv

  21. [29]

    Tra-moe: Learning trajectory prediction model from multiple domains for adaptive policy conditioning,

    Jiange Yang, Haoyi Zhu, Yating Wang, et al., “Tra-moe: Learning trajectory prediction model from multiple domains for adaptive policy conditioning,” inProceedings of the Com- puter Vision and Pattern Recognition Conference, 2025, pp. 6960–6970

    2025

  22. [30]

    Cot-vla: Visual chain-of-thought reasoning for vision-language-action models,

    Qingqing Zhao, Yao Lu, Moo Jin Kim, et al., “Cot-vla: Visual chain-of-thought reasoning for vision-language-action models,” inProceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 1702–1713

    2025

  23. [31]

    OTTER: A vision-language-action model with text-aware visual feature extraction,

    Huang Huang, Fangchen Liu, Letian Fu, et al., “OTTER: A vision-language-action model with text-aware visual feature extraction,” inF orty-second International Conference on Ma- chine Learning, 2025

    2025

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.