The study evaluated 281 imitation‑learning policies across seven families in 1,700 real‑robot trials of RJ45 connector grasping and insertion with a five‑fingered hand. Controlled ablations revealed that observation and action representations—such as ROI cropping, robot‑base framing, and point‑encoder normalization for point‑cloud policies, and absolute action prediction for vision‑language‑action policies—drastically improved success rates. Among 67 offline metrics, only five measures of predicted‑motion smoothness and directness correlated reliably with real‑world success, indicating that offline metrics can help shortlist policies but cannot replace actual robot testing.
By Ammar N. Abbas, Aayush Jain, Prashanth Viswanath, Nivesh Gadipudi, Giulia Vilone, Jim Gibbons
The paper introduces 2AM, a system that separates memory and action execution in long‑horizon robot manipulation. 2AM stores task memory exclusively in a multimodal Agent, while a single RGB‑based, stateless Action Model performs motion based on language and optional 2D hints. On the LIBERO‑Mem benchmark, 2AM achieves 76.3% average completion—over 61 points higher than the best baseline—demonstrating that agent‑side memory and precise steering of the Action Model can substantially improve performance.
By Yutong Hu, Fengjiao Chen, Xuezhi Cao, Renaud Detry
arXiv:2607. 09825v1 Announce Type: cross Abstract: Robotic manipulation policies rely on pre-trained vision models that give either a global scene embedding or a dense patch grid.
By Yi Li (TU Darmstadt), Alexandre Chapin (LIRIS), Liming Chen (LIRIS), Jan Peters (TU Darmstadt), Alap Kshirsagar (IIT Delhi, ADU)
Universal Manipulation Interface (UMI) enables scalable real-world robot data collection without hardware-specific teleoperation, yet leveraging UMI data to train large-scale Vision-Language-Action (VLA) models remains fundamentally challenging. We identify two critical mismatches: wrist-mounted fisheye views, with severe radial distortion and local gripper-centric perspectives, are out-of-distribution for pretrained VLMs; and human-collected trajectories frequently violate kinematic limits, incur collisions, or exceed controller bandwidth, teaching VLA policies physically infeasible actions.
arXiv:2609.05588v3 Announce Type: replace-cross
Abstract: World-action models (WAM) predict future states to guide robot actions, enabling learning from both action-free video and action-labeled inte...
By Renhang Liu, Wenzhi Zhao, Zhuo Yang, Liliang Chen, Pengfei Zhou, Shengcong Chen, Guanghui Ren, Youlun Peng, Rongjun Jin, Nan Wang, Sukai Wang, Xindong He, Jinyuan Feng, Ziyu Xiong, Linqing Zhong, Yifei Wei, Feng Han, Long Zhang, Da Huang, Nanshu Zhao, Chenghao Yin, Mo Wu, Zhaodong Yan, Kongtao Hu, Yuxiang Yan, Aogelijiang Niyazi, Yu Fang, Jia Zeng, Lizhu Meng, Daizhen Lv, Haoyu Cao, Zhiwen Hou, Lianjin Ye, Yuehan Niu, Zhikai Cai, Xuan Hu, Hui Min, Xiongfeng Cai, Yue Liao, Jing Wu, Soujanya Poria, Ye Li, Sanping Zhou, Maoqing Yao, AgiBot Research Team
arXiv:2606. 04708v1 Announce Type: cross Abstract: Universal Manipulation Interface (UMI) enables scalable real-world robot data collection without hardware-specific teleoperation, yet leveraging UMI data to train large-scale Vision-Language-Action (VLA) models remains fundamentally challenging.
By Siyuan Yang, Linzheng Guo, Ouyang Lu, Zhaxizhuoma, Daoran Zhang, Xinmiao Wang, Ting Xiao, Fangzheng Yan, Zhijun Chen, Yan Ding, Chao Yu, Chenjia Bai, Xuelong Li
BiDex-100 is a simulation benchmark for bimanual dexterous manipulation, featuring 106 tasks across five capability axes—generalization, instruction following, limited demonstrations, long horizon, and memory. Each task includes an out‑of‑domain variant to test robustness, and the benchmark supports two hand embodiments with 8,080 teleoperated demonstrations totaling 6.25 million frames. The authors evaluate seven recent Vision‑Language‑Action models under a unified protocol, providing baseline performance for capability, robustness, and data efficiency.
By Yong Wang, Kecheng Zheng, Xing Zhu, Chao Xu, Yujun Shen, Shengze Cai
AFFORD2ACT investigates whether robot manipulation policies can rely on a very compact visual representation instead of dense images. Using an affordance model to select up to 15 object keypoints and 4 end‑effector keypoints, the policy receives only 19 2D points (38 coordinates). In six real‑robot tasks, this sparse keypoint input outperforms RGB, RGB‑D, and dense‑affordance baselines, maintains performance with fewer demonstrations, and achieves 82% success on held‑out data, though it depends on accurate affordance localization and keypoint tracking.
By Anukriti Singh, Kasra Torshizi, Khuzema Habib, Kelin Yu, Ruohan Gao, Pratap Tokekar
The paper introduces 2AM, a system that keeps task memory solely within a multimodal Agent while using a single RGB‑based, stateless Action Model to execute motions. By compiling interaction history into subtask language and optional 2D grasp/place/move hints, the Agent steers the Action Model, which is trained to tolerate imperfect guidance through dropout, noise, and jitter. On the LIBERO‑Mem benchmark, 2AM achieves 76.3% average completion without depth, geometry, or planners, vastly outperforming the best baseline.
arXiv:2602. 09580v4 Announce Type: replace-cross Abstract: Real-world fine-tuning of dexterous manipulation policies remains challenging due to limited real-world interaction budgets and highly multimodal action distributions.
By Chenyu Yang, Denis Tarasov, Davide Liconti, Romain Guntz, Hehui Zheng, Robert K. Katzschmann
arXiv:2609.38059v1 Announce Type: cross
Abstract: Real-world robot learning is constrained by the cost of collecting experience and evaluating candidate behaviors. Video generation models offer a sca...
By Shenghe Zheng, Wenbo Li, Jiyao Zhang, Bin Xia, Haoyang Huang, Nan Duan, Jiaya Jia
arXiv:2605. 30226v2 Announce Type: replace-cross Abstract: Vision-Language-Action (VLA) models have emerged as a promising paradigm for grounding visual-language understanding into real-world robotic manipulation.
By Zhongxi Chen, Yifan Han, Yanming Shao, Huanming Liu, Congsheng Xu, Xiaoyu Chen, Yao Mu, Wenzhao Lian