The paper introduces WholeBodyMem, an object-guided perceptual memory for humanoid loco‑manipulation that records key frames of object events before the robot acts on its goal. Using SAM3 to track task objects, the system selects causally relevant frames within an adaptive budget and feeds them to an action expert via cross‑attention, along with tokens encoding appearance, geometry, and time. The authors also present HuMeGo, a benchmark with simulated and real‑world Unitree G1 tasks, showing that WholeBodyMem significantly improves success rates over memoryless and uniform‑memory baselines—raising policy success from 29.1% to 72.6% in simulation and from 36.7% to 88.3% on the real robot, while also boosting standard SIMPLE tasks from 70.2% to 86.9%.
arXiv:2609.36595v1 Announce Type: cross Abstract: Visual-memory systems commonly retain or compress past observations. Robot control additionally requires interaction-derived state that no individual...
arXiv:2603. 24576v2 Announce Type: replace-cross Abstract: Robots often observe information that determines a future action long before that action is executed.
HWAM (Humanoid World Action Model) jointly generates reference actions and the robot’s post‑execution proprioceptive state, bridging the action–execution gap in hierarchical humanoid systems. It integrates three conditional paths—policy denoising, forward dynamics modeling, and inverse dynamics modeling—to connect policy references, realized motion, and visual outcomes. On the LimX OLI humanoid, HWAM outperforms baselines, achieving a 70.6% success rate on Candy Picking versus 43.3% for Fast‑WAM.
The paper introduces HWAM, a Humanoid World Action Model that jointly generates reference actions and the robot’s post‑execution proprioceptive state, thereby closing the action‑execution gap in hierarchical humanoid systems. HWAM is trained via three conditional paths—policy denoising, forward dynamics modeling, and inverse dynamics modeling—to align policy outputs, executed motion, and visual outcomes. Experiments on the LimX OLI humanoid show HWAM achieving a 70.6% success rate on Candy Picking, outperforming Fast‑WAM’s 43.3%.
arXiv:2609.22684v1 Announce Type: cross Abstract: Memory-dependent robotic manipulation often requires later actions to use information from earlier interactions. Existing vision-language-action (VLA...
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.
The paper introduces RePlanMem-VLA, a selective visual memory system for Vision‑Language‑Action models that uses a policy’s own plan revisions as an online memory‑write signal. By detecting unusually large disagreements between temporally aligned action predictions, the method identifies task‑relevant events and updates a bounded visual memory while suppressing redundant frames. Experiments on MIKASA‑Robo‑VLA tasks and real‑robot replay data show that RePlanMem‑VLA retains essential observations in every episode and achieves high ROC‑AUC scores for event detection.
The paper introduces Latent Transition Memory (LTM), a method that stores interaction history as compact latent transitions between consecutive observations for vision‑language‑action models. Implemented as LTM‑π on the π₀.₅ architecture, it uses a frozen latent action encoder, a lightweight adapter, and causal KV caching to enable incremental inference without external reasoning models or memory‑specific annotations. Experiments on RoboMME and LIBERO‑Mem show that LTM‑π achieves state‑of‑the‑art success rates (58.54% and 80.5% respectively), and scaling the dataset or applying a 200‑token memory budget still yields high performance, demonstrating the method’s scalability and efficiency in real‑world robotic tasks.
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.
VioLA is a generalist humanoid control policy that predicts body and hand motion latents instead of joint commands, enabling a pretrained controller to execute these latents on a robot. By aligning human and robot motion in shared latent spaces, the policy can be trained on a large dataset of 140.6 million frames, 93.2% of which are human demonstrations. As a result, VioLA can follow locomotion and manipulation instructions on a real robot zero‑shot, achieving 100% locomotion success and 88.6% manipulation success without task‑specific fine‑tuning.
arXiv:2609.22538v1 Announce Type: cross Abstract: Large language model (LLM) planners can decompose natural-language instructions and select reusable robot skills, but choosing the correct skill does...
UniWALM is a unified world-action-language model that jointly learns video, action, and language in a single shared representation, using a mixture-of-transformers backbone with diffusion heads that co-denoise across modalities. By generating language as part of the same pass as action, the model closes a language‑as‑memory loop without an external planner, and introduces Flow Matching with Correction to maintain reliable action streams. Experiments show UniWALM outperforms baselines on compositional pick‑and‑place tasks and memory‑dependent benchmarks, and dramatically improves success on a real‑world Find‑the‑Item task when using generated textual memory.