pith. machine review for the scientific record. sign in

arxiv: 2604.20246 · v1 · submitted 2026-04-22 · 💻 cs.RO · cs.AI

Recognition: unknown

Cortex 2.0: Grounding World Models in Real-World Industrial Deployment

Adriana Aida , Walid Amer , Katarina Bankovic , Dhruv Behl , Fabian Busch , Annie Bhalla , Minh Duong , Florian Gienger
show 20 more authors
Rohan Godse Denis Grachev Ralf Gulde Elisa Hagensieker Junpeng Hu Shivam Joshi Tobias Knobloch Likith Kumar Damien LaRocque Keerthana Lokesh Omar Moured Khiem Nguyen Christian Preyss Ranjith Sriganesan Vikram Singh Carsten Sponner Anh Tong Dominik Tuscher Marc Tuscher Pavan Upputuri
Authors on Pith no claims yet

Pith reviewed 2026-05-10 00:46 UTC · model grok-4.3

classification 💻 cs.RO cs.AI
keywords world modelsrobotic manipulationvision-language-actiontrajectory planningindustrial roboticslatent spacelong-horizon tasksplan-and-act
0
0 comments X

The pith

Cortex 2.0 shifts from reactive vision-language-action control to generating and scoring candidate trajectories in visual latent space before acting.

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

The paper introduces Cortex 2.0 as a system that plans ahead by creating multiple possible future sequences of robot actions inside a compressed visual representation, then picks the sequence expected to succeed and finish efficiently. This replaces the standard method where models pick only the immediate next action from the current camera image. The authors demonstrate the approach on physical single-arm and dual-arm robots performing pick-and-place, sorting, and unpacking tasks in messy industrial conditions with clutter and occlusions. The planned system beats current leading reactive models on every task. If the results hold, world-model planning offers a route to dependable long-horizon robot behavior outside controlled lab settings.

Core claim

Cortex 2.0 generates candidate future trajectories in visual latent space, scores them for expected success and efficiency, then commits only to the highest-scoring candidate. This plan-and-act approach allows the system to outperform state-of-the-art reactive Vision-Language-Action models across single-arm and dual-arm platforms on tasks including pick and place, item and trash sorting, screw sorting, and shoebox unpacking, particularly in unstructured environments with heavy clutter, frequent occlusions, and contact-rich manipulation.

What carries the argument

Generation and scoring of candidate trajectories in visual latent space to select the best plan before acting.

Load-bearing premise

That trajectories scored highly in the visual latent space will translate to successful and efficient real-world executions on physical robots amid changing object arrangements.

What would settle it

A head-to-head test on a fresh industrial task with new object distributions and clutter patterns where the success rate of trajectories chosen by Cortex 2.0 is not higher than that of a reactive baseline.

Figures

Figures reproduced from arXiv: 2604.20246 by Adriana Aida, Anh Tong, Annie Bhalla, Carsten Sponner, Christian Preyss, Damien LaRocque, Denis Grachev, Dhruv Behl, Dominik Tuscher, Elisa Hagensieker, Fabian Busch, Florian Gienger, Junpeng Hu, Katarina Bankovic, Keerthana Lokesh, Khiem Nguyen, Likith Kumar, Marc Tuscher, Minh Duong, Omar Moured, Pavan Upputuri, Ralf Gulde, Ranjith Sriganesan, Rohan Godse, Shivam Joshi, Tobias Knobloch, Vikram Singh, Walid Amer.

Figure 1
Figure 1. Figure 1: Overview of Cortex 2.0 Abstract Industrial robotic manipulation demands reliable long-horizon execution across embodiments, tasks, and changing object distributions. While Vision-Language￾Action models have demonstrated strong generalization, they remain fundamentally reactive. By optimizing the next action given the current observation without evaluating potential futures, they are brittle to the compound… view at source ↗
Figure 2
Figure 2. Figure 2: Cortex 2.0 Architecture We consider a partially observed control problem with discrete time t = 1,...,T. At each step, the robot receives multimodal observations ot = (I rgb t , rt, ft, lt), (1) where I rgb t is the RGB image from wrist cameras, rt is the robot state, ft the force feedback of the end-effector, and lt the embedding of the task instruction. The observation is encoded into a visual latent zt … view at source ↗
Figure 3
Figure 3. Figure 3: PRO scores k candidate rollouts via the composite score Sj (Eq. 11). The loss landscape shows all candidate trajectories (top); PRO selects the highest-scoring rollout τ ∗ (bottom). The PRO heads are trained on real executed trajectories from deployment data, where ground-truth outcomes are available, and applied at inference time to imagined latents from the world model. PRO is pretrained and kept frozen … view at source ↗
Figure 4
Figure 4. Figure 4: World Model Rollout Generation. Each column of the [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Real Production Robotic Data. 9 [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Open Source Datasets. (a) Pick and Place - Single Arm (b) Sorting and Packaging (c) Pick and Packaging (d) Transport [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Synthetic Isaac Sim data – Dual Robotic Arms. [PITH_FULL_IMAGE:figures/full_fig_p010_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Cortex 2.0 Performance against Number of Rollouts [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Subtasks of Single-Arm Pick-and-Place. The robot detects best item to pick, transitions to [PITH_FULL_IMAGE:figures/full_fig_p013_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Subtasks Sorting Items and Trash. The robot must detect which items are trash and which [PITH_FULL_IMAGE:figures/full_fig_p013_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Subtasks of Sorting Screws. All screws are sorted into the according compartments. [PITH_FULL_IMAGE:figures/full_fig_p013_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Subtasks of Shoebox Unpacking. The shoebox is opened, packaging paper is removed [PITH_FULL_IMAGE:figures/full_fig_p014_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Performance comparison across all four benchmark tasks. [PITH_FULL_IMAGE:figures/full_fig_p016_13.png] view at source ↗
read the original abstract

Industrial robotic manipulation demands reliable long-horizon execution across embodiments, tasks, and changing object distributions. While Vision-Language-Action models have demonstrated strong generalization, they remain fundamentally reactive. By optimizing the next action given the current observation without evaluating potential futures, they are brittle to the compounding failure modes of long-horizon tasks. Cortex 2.0 shifts from reactive control to plan-and-act by generating candidate future trajectories in visual latent space, scoring them for expected success and efficiency, then committing only to the highest-scoring candidate. We evaluate Cortex 2.0 on a single-arm and dual-arm manipulation platform across four tasks of increasing complexity: pick and place, item and trash sorting, screw sorting, and shoebox unpacking. Cortex 2.0 consistently outperforms state-of-the-art Vision-Language-Action baselines, achieving the best results across all tasks. The system remains reliable in unstructured environments characterized by heavy clutter, frequent occlusions, and contact-rich manipulation, where reactive policies fail. These results demonstrate that world-model-based planning can operate reliably in complex industrial environments.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper introduces Cortex 2.0, a world-model-based planning system for industrial robotic manipulation. It generates candidate future trajectories in visual latent space, scores them for expected success and efficiency, and executes only the highest-scoring candidate. The authors evaluate the system on single-arm and dual-arm platforms across four tasks of increasing complexity (pick and place, item and trash sorting, screw sorting, shoebox unpacking) and claim consistent outperformance over state-of-the-art Vision-Language-Action baselines, with particular reliability in unstructured environments featuring heavy clutter, frequent occlusions, and contact-rich manipulation.

Significance. If the results hold, the work would demonstrate that latent-space world models can be grounded for reliable long-horizon planning in real industrial settings, providing a concrete alternative to reactive VLA policies that fail under compounding errors. This would strengthen the case for deploying planning-based approaches in contact-rich, partially observable domains.

major comments (2)
  1. [Evaluation and Experiments (results on the four tasks)] The central claim that Cortex 2.0 'consistently outperforms' VLA baselines and 'remains reliable' in unstructured settings rests on the unverified assumption that trajectories scored in visual latent space accurately rank real-world success and efficiency. In heavy clutter, occlusions, and contact-rich manipulation, visual latents typically discard precise 3D geometry, mass, friction, and partial-observability information; without a dedicated validation (e.g., correlation between latent scores and physical outcomes or failure-mode analysis), outperformance cannot be attributed to the world-model planner.
  2. [Abstract and Results sections] No quantitative results, error bars, baseline implementations, or scoring-function details are supplied to support the outperformance claim across the four tasks. This absence prevents assessment of whether the reported gains are statistically meaningful or task-specific.
minor comments (1)
  1. [Abstract] The abstract would be clearer if it briefly defined the success and efficiency metrics used for trajectory scoring.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their insightful comments on our work. We address each of the major comments point-by-point below, providing clarifications and indicating revisions to the manuscript.

read point-by-point responses

  1. Referee: [Evaluation and Experiments (results on the four tasks)] The central claim that Cortex 2.0 'consistently outperforms' VLA baselines and 'remains reliable' in unstructured settings rests on the unverified assumption that trajectories scored in visual latent space accurately rank real-world success and efficiency. In heavy clutter, occlusions, and contact-rich manipulation, visual latents typically discard precise 3D geometry, mass, friction, and partial-observability information; without a dedicated validation (e.g., correlation between latent scores and physical outcomes or failure-mode analysis), outperformance cannot be attributed to the world-model planner.

    Authors: We agree that explicit validation of the latent scoring mechanism is important for attributing the observed outperformance to the world-model planner. The current manuscript demonstrates this through end-to-end task performance, but to address the concern directly, we will include in the revised version a quantitative correlation study between the predicted scores and measured success/efficiency metrics, as well as an analysis of failure cases where high-scoring trajectories led to suboptimal outcomes. This will clarify the grounding of the visual latents for the specific industrial tasks. revision: yes

  2. Referee: [Abstract and Results sections] No quantitative results, error bars, baseline implementations, or scoring-function details are supplied to support the outperformance claim across the four tasks. This absence prevents assessment of whether the reported gains are statistically meaningful or task-specific.

    Authors: The manuscript does present quantitative success rates for Cortex 2.0 and the VLA baselines across the four tasks in the results section. However, we acknowledge that error bars, detailed baseline specifications, and the precise scoring function equations are not fully detailed. We will revise the abstract to include key quantitative highlights, add error bars to the performance tables and figures, specify the baseline implementations (including model versions and training details), and expand the methods section with the complete scoring function formulation. These additions will enable a full statistical evaluation of the results. revision: yes

Circularity Check

0 steps flagged

No circularity in claimed derivation or predictions

full rationale

The paper presents an empirical robotic system (Cortex 2.0) that generates and scores trajectories in visual latent space before execution. No equations, first-principles derivations, or 'predictions' of new quantities from fitted parameters appear in the provided text. Performance claims rest on experimental comparisons across tasks rather than any self-referential reduction or self-citation chain that would force the result by construction. The central assumption (latent scores predict real-world outcomes) is an empirical hypothesis open to falsification, not a definitional tautology.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review yields no explicit free parameters, axioms, or invented entities; the latent-space world model and scoring mechanism are described at a high level without stated assumptions or fitted constants.

pith-pipeline@v0.9.0 · 5603 in / 1164 out tokens · 42501 ms · 2026-05-10T00:46:51.772218+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

42 extracted references · 26 canonical work pages · 18 internal anchors

  1. [1]

    $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control

    Kevin Black, Noah Brown, Danny Driess, Adnan Esmail, Michael Equi, Chelsea Finn, Niccolo Fusai, Lachy Groom, Karol Hausman, Brian Ichter, et al. π0: A Vision-Language-Action Flow Model for General Robot Control.arXiv preprint arXiv:2410.24164, 2024

    work page internal anchor Pith review arXiv 2024

  2. [2]

    $\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization

    Physical Intelligence, Kevin Black, Noah Brown, James Darpinian, Karan Dhabalia, Danny Driess, Adnan Esmail, Michael Equi, Chelsea Finn, Niccolo Fusai, et al. π0.5: A Vision- Language-Action Model with Open-World Generalization.arXiv preprint arXiv:2504.16054, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  3. [3]

    RT-1: Robotics Transformer for Real-World Control at Scale

    Anthony Brohan, Noah Brown, Justice Carbajal, Yevgen Chebotar, Joseph Dabis, Chelsea Finn, Keerthana Gopalakrishnan, Karol Hausman, Alex Herzog, Jasmine Hsu, et al. RT-1: Robotics Transformer for Real-World Control at Scale.arXiv preprint arXiv:2212.06817, 2022

    work page internal anchor Pith review arXiv 2022

  4. [4]

    RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control

    Brianna Zitkovich, Tianhe Yu, Sichun Xu, Peng Xu, Ted Xiao, Fei Xia, Jialin Wu, Paul Wohlhart, Stefan Welker, Ayzaan Wahid, et al. RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control. InConference on Robot Learning, pages 2165–2183. PMLR, 2023

    2023

  5. [5]

    Danny Driess, Fei Xia, Mehdi S. M. Sajjadi, Corey Lynch, Aakanksha Chowdhery, Brian Ichter, Ayzaan Wahid, Jonathan Tompson, Quan Vuong, Tianhe Yu, et al. PaLM-E: An Embodied Multimodal Language Model.arXiv preprint arXiv:2303.03378, 2023

    work page internal anchor Pith review arXiv 2023

  6. [6]

    Open X- 17 Embodiment: Robotic Learning Datasets and RT-X Models

    Abby O’Neill, Abdul Rehman, Abhiram Maddukuri, Abhishek Gupta, Abhishek Padalkar, Abraham Lee, Acorn Pooley, Agrim Gupta, Ajay Mandlekar, Ajinkya Jain, et al. Open X- 17 Embodiment: Robotic Learning Datasets and RT-X Models. In2024 IEEE International Conference on Robotics and Automation (ICRA), pages 6892–6903. IEEE, 2024

    2024

  7. [7]

    Bridge Data: Boosting Generalization of Robotic Skills with Cross-Domain Datasets

    Frederik Ebert, Yanlai Yang, Karl Schmeckpeper, Bernadette Bucher, Georgios Georgakis, Kostas Daniilidis, Chelsea Finn, and Sergey Levine. Bridge Data: Boosting Generalization of Robotic Skills with Cross-Domain Datasets.arXiv preprint arXiv:2109.13396, 2021

    work page internal anchor Pith review arXiv 2021

  8. [8]

    Learning interactive real-world simulators.arXiv preprint arXiv:2310.06114, 1(2):6, 2023

    Mengjiao Yang, Yilun Du, Kamyar Ghasemipour, Jonathan Tompson, Dale Schuurmans, and Pieter Abbeel. Learning Interactive Real-World Simulators.arXiv preprint arXiv:2310.06114, 1(2):6, 2023

    work page arXiv 2023

  9. [9]

    Cosmos World Foundation Model Platform for Physical AI

    Niket Agarwal, Arslan Ali, Maciej Bala, Yogesh Balaji, Erik Barker, Tiffany Cai, Prithvijit Chattopadhyay, Yongxin Chen, Yin Cui, Yifan Ding, et al. Cosmos World Foundation Model Platform for Physical AI.arXiv preprint arXiv:2501.03575, 2025

    work page internal anchor Pith review arXiv 2025

  10. [10]

    Octo: An Open-Source Generalist Robot Policy

    Oier Mees, Dibya Ghosh, Karl Pertsch, Kevin Black, Homer Rich Walke, Sudeep Dasari, Joey Hejna, Tobias Kreiman, Charles Xu, Jianlan Luo, et al. Octo: An Open-Source Generalist Robot Policy. InFirst Workshop on Vision-Language Models for Navigation and Manipulation at ICRA 2024, 2024

    2024

  11. [11]

    OpenVLA: An Open-Source Vision-Language-Action Model

    Moo Jin Kim, Karl Pertsch, Siddharth Karamcheti, Ted Xiao, Ashwin Balakrishna, Suraj Nair, Rafael Rafailov, Ethan Foster, Grace Lam, Pannag Sanketi, et al. OpenVLA: An Open-Source Vision-Language-Action Model.arXiv preprint arXiv:2406.09246, 2024

    work page internal anchor Pith review arXiv 2024

  12. [12]

    Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success

    Moo Jin Kim, Chelsea Finn, and Percy Liang. Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success.arXiv preprint arXiv:2502.19645, 2025

    work page internal anchor Pith review arXiv 2025

  13. [13]

    MolmoAct: Action Reasoning Models that can Reason in Space

    Jason Lee, Jiafei Duan, Haoquan Fang, Yuquan Deng, Shuo Liu, Boyang Li, Bohan Fang, Jieyu Zhang, Yi Ru Wang, Sangho Lee, et al. MolmoAct: Action Reasoning Models That Can Reason in Space.arXiv preprint arXiv:2508.07917, 2025

    work page internal anchor Pith review arXiv 2025

  14. [14]

    Soroush Nasiriany, Abhiram Maddukuri, Lance Zhang, Adeet Parikh, Aaron Lo, Abhishek Joshi, Ajay Man- dlekar, and Yuke Zhu

    Qi Lv, Weijie Kong, Hao Li, Jia Zeng, Zherui Qiu, Delin Qu, Haoming Song, Qizhi Chen, Xiang Deng, and Jiangmiao Pang. F1: A Vision-Language-Action Model Bridging Understanding and Generation to Actions.arXiv preprint arXiv:2509.06951, 2025

    work page arXiv 2025

  15. [15]

    FAST: Efficient Action Tokenization for Vision-Language-Action Models

    Karl Pertsch, Kyle Stachowicz, Brian Ichter, Danny Driess, Suraj Nair, Quan Vuong, Oier Mees, Chelsea Finn, and Sergey Levine. FAST: Efficient Action Tokenization for Vision-Language- Action Models.arXiv preprint arXiv:2501.09747, 2025

    work page internal anchor Pith review arXiv 2025

  16. [16]

    GR00T N1: An Open Foundation Model for Generalist Humanoid Robots

    Johan Bjorck, Fernando Castañeda, Nikita Cherniadev, Xingye Da, Runyu Ding, Linxi Fan, Yu Fang, Dieter Fox, Fengyuan Hu, Spencer Huang, et al. GR00T N1: An Open Foundation Model for Generalist Humanoid Robots.arXiv preprint arXiv:2503.14734, 2025

    work page internal anchor Pith review arXiv 2025

  17. [17]

    Cortex: Bridging Vision, Language, and Action with Discrete Plans and Tokens

    Sereact. Cortex: Bridging Vision, Language, and Action with Discrete Plans and Tokens. Sereact Technical Blog, September 2025. URL https://sereact.ai/posts/ cortex-bridging-vision-language-and-action-with-discrete-plans-and-tokens

    2025

  18. [18]

    Diffusion Policy: Visuomotor Policy Learning via Action Diffusion

    Cheng Chi, Zhenjia Xu, Siyuan Feng, Eric Cousineau, Yilun Du, Benjamin Burchfiel, Russ Tedrake, and Shuran Song. Diffusion Policy: Visuomotor Policy Learning via Action Diffusion. The International Journal of Robotics Research, 44(10-11):1684–1704, 2025

    2025

  19. [19]

    Denoising Diffusion Probabilistic Models.Advances in Neural Information Processing Systems, 33:6840–6851, 2020

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising Diffusion Probabilistic Models.Advances in Neural Information Processing Systems, 33:6840–6851, 2020

    2020

  20. [20]

    Yaron Lipman, Ricky T. Q. Chen, Heli Ben-Hamu, Maximilian Nickel, and Matt Le. Flow Matching for Generative Modeling.arXiv preprint arXiv:2210.02747, 2022

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  21. [21]

    Albergo and Eric Vanden-Eijnden

    Michael S. Albergo and Eric Vanden-Eijnden. Building Normalizing Flows with Stochastic Interpolants. InInternational Conference on Learning Representations, 2023

    2023

  22. [22]

    Boffi, and Eric Vanden-Eijnden

    Michael Albergo, Nicholas M. Boffi, and Eric Vanden-Eijnden. Stochastic Interpolants: A Unifying Framework for Flows and Diffusions.Journal of Machine Learning Research, 26 (209):1–80, 2025

    2025

  23. [23]

    arXiv preprint arXiv:2602.03310 (2026)

    Songming Liu, Bangguo Li, Kai Ma, Lingxuan Wu, Hengkai Tan, Xiao Ouyang, Hang Su, and Jun Zhu. RDT2: Exploring the Scaling Limit of UMI Data Towards Zero-Shot Cross- Embodiment Generalization.arXiv preprint arXiv:2602.03310, 2026

    work page arXiv 2026

  24. [24]

    World Models

    David Ha and Jürgen Schmidhuber. World Models.arXiv preprint arXiv:1803.10122, 2(3):440, 2018. 18

    work page internal anchor Pith review arXiv 2018

  25. [25]

    Deep Visual Foresight for Planning Robot Motion

    Chelsea Finn and Sergey Levine. Deep Visual Foresight for Planning Robot Motion. In2017 IEEE International Conference on Robotics and Automation (ICRA), pages 2786–2793. IEEE, 2017

    2017

  26. [26]

    Mastering Atari with Discrete World Models

    Danijar Hafner, Timothy Lillicrap, Mohammad Norouzi, and Jimmy Ba. Mastering Atari with Discrete World Models.arXiv preprint arXiv:2010.02193, 2020

    work page internal anchor Pith review arXiv 2010

  27. [27]

    Mastering Diverse Domains through World Models

    Danijar Hafner, Jurgis Pasukonis, Jimmy Ba, and Timothy Lillicrap. Mastering Diverse Domains Through World Models.arXiv preprint arXiv:2301.04104, 2023

    work page internal anchor Pith review arXiv 2023

  28. [28]

    IRASim: A Fine-Grained World Model for Robot Manipulation

    Fangqi Zhu, Hongtao Wu, Song Guo, Yuxiao Liu, Chilam Cheang, and Tao Kong. IRASim: A Fine-Grained World Model for Robot Manipulation. InProceedings of the IEEE/CVF International Conference on Computer Vision, pages 9834–9844, 2025

    2025

  29. [29]

    Strengthening Generative Robot Policies through Predictive World Modeling, May 2025

    Han Qi et al. Strengthening Generative Robot Policies Through Predictive World Modeling. arXiv preprint arXiv:2502.00622, 2025

    work page arXiv 2025

  30. [30]

    GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot Manipulation

    Chi-Lam Cheang, Guangzeng Chen, Ya Jing, Tao Kong, Hang Li, Yifeng Li, Yuxiao Liu, Hongtao Wu, Jiafeng Xu, Yichu Yang, et al. GR-2: A Generative Video-Language-Action Model with Web-Scale Knowledge for Robot Manipulation.arXiv preprint arXiv:2410.06158, 2024

    work page internal anchor Pith review arXiv 2024

  31. [31]

    V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning

    Mido Assran, Adrien Bardes, David Fan, Quentin Garrido, Russell Howes, Matthew Muckley, Ammar Rizvi, Claire Roberts, Koustuv Sinha, Artem Zholus, et al. V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning.arXiv preprint arXiv:2506.09985, 2025

    work page internal anchor Pith review arXiv 2025

  32. [32]

    Robotic world model: A neural network simulator for robust policy optimization in robotics.arXiv preprint arXiv:2501.10100, 2025

    Chenhao Li et al. Robotic World Model: A Neural Network Simulator for Robust Policy Optimization in the Real World.arXiv preprint arXiv:2501.10100, 2025

    work page arXiv 2025

  33. [33]

    Adelson, and Sergey Levine

    Roberto Calandra, Andrew Owens, Dinesh Jayaraman, Justin Lin, Wenzhen Yuan, Jitendra Malik, Edward H. Adelson, and Sergey Levine. More Than a Feeling: Learning to Grasp and Regrasp Using Vision and Touch.IEEE Robotics and Automation Letters, 2018

    2018

  34. [34]

    Lee, Yuke Zhu, Peter Zachares, Matthew Tan, Krishnan Srinivasan, Silvio Savarese, Li Fei-Fei, Animesh Garg, and Jeannette Bohg

    Michelle A. Lee, Yuke Zhu, Peter Zachares, Matthew Tan, Krishnan Srinivasan, Silvio Savarese, Li Fei-Fei, Animesh Garg, and Jeannette Bohg. Making Sense of Vision and Touch: Learning Multimodal Representations for Contact-Rich Tasks.IEEE Transactions on Robotics, 36 (3):582–596, June 2020. ISSN 1941-0468. doi: 10.1109/TRO.2019.2959445. URL http: //dx.doi....

    work page doi:10.1109/tro.2019.2959445 2020

  35. [35]

    Zhao, Quan Vuong, Chongyi Zheng, Philippe Hansen- Estruch, Andre Wang He, Vivek Myers, Moo Jin Kim, Max Du, et al

    Homer Rich Walke, Kevin Black, Tony Z. Zhao, Quan Vuong, Chongyi Zheng, Philippe Hansen- Estruch, Andre Wang He, Vivek Myers, Moo Jin Kim, Max Du, et al. BridgeData V2: A Dataset for Robot Learning at Scale. InConference on Robot Learning, pages 1723–1736. PMLR, 2023

    2023

  36. [36]

    DROID: A Large-Scale In-The-Wild Robot Manipu- lation Dataset, 2024

    Alexander Khazatsky, Karl Pertsch, and ... DROID: A Large-Scale In-The-Wild Robot Manipu- lation Dataset, 2024

    2024

  37. [37]

    AgiBot World Colosseo: A Large-scale Manipulation Platform for Scalable and Intelligent Embodied Systems

    Qingwen Bu, Jisong Cai, Li Chen, Xiuqi Cui, Yan Ding, Siyuan Feng, Shenyuan Gao, Xindong He, Xu Huang, Shu Jiang, et al. AgiBot World Colosseo: A Large-Scale Manipulation Platform for Scalable and Intelligent Embodied Systems.arXiv preprint arXiv:2503.06669, 2025

    work page internal anchor Pith review arXiv 2025

  38. [38]

    Schoellig, and Timothy D

    Jianlan Luo, Charles Xu, Xinyang Geng, Gilbert Feng, Kuan Fang, Liam Tan, Stefan Schaal, and Sergey Levine. Multistage Cable Routing Through Hierarchical Imitation Learning.IEEE Transactions on Robotics, 40:1476–1491, 2024. ISSN 1941-0468. doi: 10.1109/TRO.2024. 3353075. URLhttp://dx.doi.org/10.1109/TRO.2024.3353075

    work page doi:10.1109/tro.2024 2024

  39. [39]

    Zhao, Vikash Kumar, Sergey Levine, and Chelsea Finn

    Tony Z. Zhao, Vikash Kumar, Sergey Levine, and Chelsea Finn. Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware. InRobotics: Science and Systems, 2023

    2023

  40. [40]

    Knowledge insulating vision-language-action models: Train fast, run fast, generalize better.arXiv preprint arXiv:2505.23705, 2025

    Danny Driess, Jost Tobias Springenberg, Brian Ichter, Lili Yu, Adrian Li-Bell, Karl Pertsch, Allen Z. Ren, Homer Walke, Quan Vuong, Lucy Xiaoyang Shi, et al. Knowledge Insulating Vision-Language-Action Models: Train Fast, Run Fast, Generalize Better.arXiv preprint arXiv:2505.23705, 2025

    work page arXiv 2025

  41. [41]

    RoboCasa: Large-Scale Simulation of Everyday Tasks for Generalist Robots

    Soroush Nasiriany, Abhiram Maddukuri, Lance Zhang, Adeet Parikh, Aaron Lo, Abhishek Joshi, Ajay Mandlekar, and Yuke Zhu. RoboCasa: Large-Scale Simulation of Everyday Tasks for Generalist Robots. InRobotics: Science and Systems (RSS), 2024. 19

    2024

  42. [42]

    LeRobot: State-of-the-Art Machine Learning for Real-World Robotics in PyTorch

    Remi Cadene, Simon Alibert, Alexander Soare, Quentin Gallouedec, Adil Zouitine, Steven Palma, Pepijn Kooijmans, Michel Aractingi, Mustafa Shukor, Dana Aubakirova, Martino Russi, Francesco Capuano, Caroline Pascale, Jade Choghari, Jess Moss, and Thomas Wolf. LeRobot: State-of-the-Art Machine Learning for Real-World Robotics in PyTorch. https: //github.com/...

    2024