The paper explores whether inexpensive, procedurally generated synthetic data can provide useful priors for robot learning models before any real-world data is observed. By creating three data generation pipelines that capture mathematical abstractions of robotic control, the authors pretrain a Transformer and evaluate its zero‑shot action prediction on unseen physical robot tasks. The results show that the model achieves non‑trivial performance via in‑context learning, with specialized control priors outperforming generic causal priors, though a compute‑matched real‑robot baseline still excels on its own training datasets.
Recent progress in large-scale imitation learning for robot manipulation has been driven by leveraging datasets across a wide range of robot embodiments. However, achieving significant cross-embodiment transfer is often still challenging.
InternW0-Δ is a unified World Action Model that integrates pretrained visual dynamics, scene semantics, 4D geometry, and motion priors within a Mixture-of-Transformers framework to generate robot actions. It leverages a frozen VLM for semantic guidance, a 4D foundation model for geometric priors, and introduces Causal Imprint to learn future-relevant scene changes without future-video rollout. The model is pretrained on a newly curated 20K‑hour heterogeneous corpus of robot and human demonstrations, achieving superior performance on simulation benchmarks and real‑robot platforms.
arXiv:2607. 02466v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models are fundamentally bottlenecked by the scarcity of expert demonstrations -- triplets of observations, instructions, and actions that are costly to collect at scale.
arXiv:2607. 27549v1 Announce Type: cross Abstract: Recent progress in large-scale imitation learning for robot manipulation has been driven by leveraging datasets across a wide range of robot embodiments.
Zero-shot cross-task generalization, where a policy must execute manipulation tasks never seen during training, remains a central challenge in robot learning. In large language models, a novel task ca...
arXiv:2608. 10600v1 Announce Type: cross Abstract: Skill abstraction---the process of learning reusable and temporally extended behaviors---has emerged as a key paradigm for improving sample efficiency and generalization in robot learning.
Scaling robot learning requires large-scale, diverse demonstrations, yet real-world data collection via teleoperation remains prohibitively expensive and time-consuming. While video diffusion models offer a promising avenue for data scaling, existing generative approaches are often limited to superficial visual augmentation, or suffer from embodiment hallucinations that yield physically infeasible motions.
Zero-WAM introduces a causal video-action model that enables robots to perform unseen manipulation tasks by following in-context human video guidance. The authors create HumanGen, a dataset of 74.2K human-robot ICL pairs across 8.6K tasks, and propose an in-context future chunk prediction objective to prevent shortcut learning. In simulation, Zero-WAM attains a 47.0% success rate on seven unseen tasks, outperforming the best video-action baseline by 29.5 percentage points, and demonstrates real‑world generalization to complex, long‑horizon, and fine‑grained tasks.
World models endow perceptual systems with the ability to predict how scenes evolve under interaction. They are most beneficial when trained on diverse volumes of data, to instill a rich prior into do...
Robot learning must produce policies that generalize to new combinations of constraints, teammates, and environments. To achieve this, we must structurally factor the policy, which is a choice that dictates what generalizes, what requires retraining, and what remains entangled.
Action-conditioned world models have become an important foundation for embodied prediction, planning, and synthetic data generation, but their errors under new task and scene distributions are often...
Synthesizing realistic Human-Object Interactions (HOI) is critical for creating embodied avatars and functional virtual environments. However, current data-driven approaches primarily rely on motion capture datasets, which are expensive to scale and limited in functional diversity.